I haven't tried Just Heroes in actual play, but it has some really nifty ideas. And since it is still in development, the pdfs are readily available.

Disregarding specific examples, I'm thinking there are two variables that determines the difficulty of this reality jumping. One I'll call distance, and one I'll call precision.

Distance is a measure of how far back the split between universes occurred. So a universe where Napoleon won the Napoleonic war is more distant than one where your opponent actually missed you with that gunshot three seconds ago.

Precision is a measure of how many "subevents" that need to line up just right, to arrive at the present that you want. So a universe where your opponent purely misses is less precise than one where he actually hits the guy behind you, which is less precise than a universe where he hits that guy right between the eyes.

More distant universes inherently requires more energy to get an equally precise present. This is because more events need to line up just right, to arrive at the same result. If your opponent shot you yesterday, and you want to end up in the exactly same situation that you are in now, except you don't have a gunshot wound (IE, he didn't hit you yesterday), you are going to need to spend a lot more energy, than if you had negated the gunshot wound three seconds after he shot you.

How difficult you want more distant places/more precise events to be, depends on the sort of campaign/setting you want to have.

With cheap distance but expensive precision, the players can easily jump to the universe where Napoleon won the war, but will have a hard time controlling what exactly is going on in that universe. For short jumps, that might mean jumping so that your opponent didn't shoot you, but instead finding yourself in a universe where you got knocked out the window instead.

With expensive distance but cheap precision, players purely be changing "local" stuff. So no traveling to alternate timelines. Instead the jumpers have a large degree of control over the events of now and a few seconds back. Impossible dodging, incredible luck and other Matrix-style stuff occurs.

With expensive distance and expensive precision, jumping more or less just becomes a reroll. You can change the immediate past, but you have no control over what you get instead.

With cheap distance and cheap precision, you get both the possibility of alternate timelines, and Matrix-style dodging and precision shots. In addition there comes the question of how much more expensive that precision becomes with distance. Can you jump to the exact same situation that you are in now, but in an alternate universe where Napoleon won the war, so that of being a cop, your opponent is now a gendarm and a member of the illuminati? Or the exact same situation in an alternate world where dinosaurs involved into dragons, so in just a sec, a dragon is going to burst in through the roof.

No matter what configuration, I'm thinking of three general moves, one dealing with distant jumps, and two dealing with precise jumps, with some tweaks depending on the exact setup that you want. Generally the moves generally results in less precision, the worse the roll.

Distant Jumping
State what kind of alternate timeline you want to end up in (one changed event in the past)
10+: You get more or less exactly what you wanted (A Franco-dominated world)
7-9: You get what you wanted, but with other major events interfering (Napoleon won, securing Europe, but Japan won the world war in 1928 and subjugated North America)
6-: What you wanted happened, but was completely negated (Napoleon won, but revolutions in the 1880s resulted in a German-dominated European Republic)

Precise Jumping(Defensive)
State one recent hard move that you would want to negate
10+: The hard move is negated, and you get a benefit (Opponent misses you, and hits his ally)
7-9: The hard move is negated (Opponent misses you)
6-: The hard move is negated, but the GM gets to make another hard move (Opponent misses you, but then knocks you out the window instead)

Precise Jumping(Offensive)
Change the result of a recent move you've made.
10+: Treat as if though you rolled 10+ on the original roll.
7-9: Treat as if though you rolled 7-9 on the original roll.
6-: The result of the move you made is negated, and the GM gets to make a hard move.

To keep the changes easy to handle, you might state that Precise Jumping can only be used immediately after the affected move / hard move. That way you won't have to retcon everything that happened between the move and the Precise Jump. A move that happened earlier would then be handled with Distant Jumping, which more easily handles such fudging.

Depending on where you want to put your Distant/Precise levers, you can apply costs to the moves, or remove them outright. Maybe you need some kind of hold to do some or all of the moves. Maybe there are penalties depending on the situation. Maybe the stat that is rolled against changes depending on situation (using some mechanism to charge up your jump device from +0, to +1, +2 and finally +3). Your Reality Collapse of course also serves that.

I also imagine that there is space for a move that could somehow express the insanity that start happening, if two jumpers start altering and counter-altering reality during a fight. That could just be handled by a load of Precise Jumping moves, but another move might be able to streamline that without losing flavor. Haven't quite figured that out yet, though.

Oh, one other note. So I'm also working on changing the names of the playsheets. I was heavily inspired by the Natures in "Night Witches", but I think using Humors or Roles for the playsheets is actually limiting in terms of expanding the game. So the list will start off with the basic doctors and nurses, but be expanded to more easily allow play for other people in the camp.

For instance:

CORPSMAN
CUTTER
DENTIST
MP
NURSE
ROLLER
SHRINK
etc.

Without actually having played, I still think that the roles are essential. When you look at MASH or Band of Brothers or whatever, they tell you who that person is in the group, much more than whether he is a nurse or a preacher or a machine gunner. Joey isn't Rifleman #3, he is the Prankster. Roles is also a group that you could maybe expand on later on.

About the Humors you might be right. But I'm still cautious about making the professions a defining part of the character. Depends a bit on the focus. I could see may moving the Humor moves over to Roles instead. Then maybe have professions, but have them be the smaller part, with a package of profession specific moves. So a character can be the Prankster Nurse or Devout Radio Operator. Most of the social/camp moves would come from the role (unless perhaps it is a social profession like preacher, psychologist, local black marketeer), while the profession would provide some professional moves (better at operations, better at fighting, able to manipulate the mission pool).

This is bases on my impression that most of the game would be about interpersonal relationships in the MASH camp (with occasional "adventures" to completment the social play). So to use the above theoretical setup, role moves for the main play in camp, and professional moves occasionally useful for the "adventures".

To put it another way, once you get past the fact that you need to speak with the supply officers to get the noise makers you need for the party, the most important thing about him isn't that he is the Supply Officer. The most important thing when you want to wrangle those supplies out of him is whether he is a Stickler, a Scrounger or maybe a Casanova. When you are a second lieutenant, once you get past the fact that the Major seriously outranks you, the most important thing isn't that he is an Officer. It's whether he is a Bully, a Mentor or, god forbid, a Prankster. At least that's how I see it.

Okay, so two topics.

Assenting: This is a sliding scale thing. On the one hand you shouldn't try to get everything your way, but on another you probably shouldn't play in a game that consistently works in a way you don't like. I was in a game recently, which seemed too focused on combat balance over player choice. I commented on it to the GM, and he made some changes, but it wasn't really enough. So I ended up leaving the game. I don't necessarily think the GM was doing anything wrong. The other players were having a great time. But it wasn't the right game for me.

So, assent if something is below your threshold. Comment on it if it is not, and try to reach a compromise, or a situation that is outright better for everyone (maybe everyone would actually prefer the second timeline). And stop commenting before you reach the asshole threshold where you are lowering the others enjoyment to raise yours. At that point you either need to decide that you didn't care that much anyway, or find a group that better meets your needs. (Which might end up being the same group playing a different campaign or rulesystem)

Timelines: I definitely prefer the second also, but there are also times for the first. But the second by default, definitely. Thief wants to use Disarm traps? Sure, okay. "But what do you do?" Another way to look at the situation (instead of fictional vs real timeline) is that we are looking at a mechanics vs fiction issue. The Alexandrian had a blog post about it just a few days ago: http://thealexandrian.net/wordpress/37976/roleplaying-games/art-of-rulings-part-3-the-fiction-mechanics-cycle

So the thief starts of by implementing mechanics, and after the check the GM comes in with the fiction. Usually I'd want the thief to provide fiction also in these cases, but in this case it might actually be hard, since the player doesn't necessarily know what would work for disarming wards in this setting. So it is up to the GM to provide the fiction, or at least help with that.

Thief: I've got a move called Disarm Traps. Can I use that?
GM: Sure! Roll +Dex
Thief: I succeed
GM: Okay, so you pull your divining rods out from your kit of thief's tools, and disrupt the magical aura around the ward for long enough that you can smudge out part of the rune without activating it.

or maybe

Thief: I've got a move called Disarm Traps. Can I use that?
GM: Yeah, sure, you know how to disarm magical runes such as these. You need to smudge out the rune, but first you need to disrupt the magical aura. You've got some gear for that in your thief's tools.
Thief: Okay, I pull out the necessary tools and try to disrupt the aura. And.. I succeed.
GM: Okay, so you manage to disrupt the magical aura around the ward, and then quickly smudge out part of the rune.

But I wouldn't respond well to "Oh you know... we'll figure that out after we see how the roll comes out." either. That's the real sin as I see it in this situation. The GM should preferably have at least an idea about what is happening in the fiction, before the roll. Otherwise he has no idea whether the thing might be possible to begin with. And if reality depends on the dice rolles, then player choice loses meaning, with everything coming down to whether you have the right moves and manage your dice rolls.

With the caveat that I've actually never been in a situation where a Battlebabe tried to use the naked option, I'd say: Naked enough that you wouldn't consider him/her even partially dressed anymore. How naked that is, probably depends upon the setting, group and maybe even situation. For some groups a bikini might work. For some underwear might be nearly naked. And in some settings/situations, only entirely naked will do.

If a player builds his or her Battlebabe on running around nearly naked, that is probably also going to be an influence, leaning towards the more dressed, bikini, part of the spectrum. But in other games it might only come up if the Battlebabe gets caught in the bath or the bedroom, entirely or almost entirely naked.

"If you trust that the medic who triaged this patient did so accurately, you can skip the Diagnose (DX) move." You should probably give a reason why Dx is needed. Because, as I read the rules right now, the safest thing to do seems to be to trust the other medic and not make a Dx roll, given how Dx has a tendency of making things worse. There is literally no benefit from rolling Dx as it stands now, and a bunch of drawbacks if you roll low (even just 7-9). Maybe move the 10+ result down to 7-9, and have a good diagnosis (10+) give a bonus to the Treatment. Playtesting will have to show the right balance, but Dx seems punitive at the moment.

The full rules look less rolley than I feared. Depending on the severity the MC wants to put into a situation, it looks like he could put in patients that only has a single clock at 4, making the surgery go by quickly unless there are complications. After playtesting, you should probably put in some examples of different configurations of patients and how they'll generally work out. Say a three day stint with two easy anonymous patients (one clock at 4), followed by a seriously injured named patient, would have the first two mostly serve to drive up stress, before getting into the hard job of saving the named patient. It might tend to reduce a patients to a bunch of numbers that just need to be reduced to 3, but that might be a feature, allowing a shock when it turns out that there is a person behind the numbers.

There does seem a bit of a lack of obvious ways for the players to strategise their way out of bad situations. In an battle you can run away, try to outmanoeuvre or try to negotiate. In the surgery it seems you can swap places with another medic, use hold (if you have it), or impose a debility. They all require a Tx roll to solve the situation, perhaps with a tiny bit of bonus. That might also be a feature, leaving a surgeon to have a bad day where he just can't help but make mistakes (bad dice luck), but I could see it sapping player enthusiasm, when they can't do anything to alleviate the bad dice rolls. The character's moves might change that, but until then..

Speaking of debility "You can now ignore this clock, but the overall clock is unaffected." First off, does this apply to all debility choices, or only Tx? Secondly, I take this to mean that the Status clock is unaffected. Or is the overall clock another clock? What is the reasoning behind not (potentially) lowering the status clock? If all other clocks are at or below 3, you are forcing the player to operate on a clock that doesn't need it. If any clocks are above 3, he'd need to operate anyway, since he can't send away the patient before all clocks are low enough.

All that being said, the conversation in surgery still seem kinda structured around the moves. I tried to run the conversation in my head, and it kinda runs make move -> explain result -> make move -> explain result -> make move. Where an AW game would go do something->make move -> explain result -> do something -> make move. The next move for a surgery always seems predestined. The player might wrap some language around it, but he is just saying stuff to justify the move that has to come. He isn't doing something, and then making a move based on what he does. But I might just be running the conversation wrong in my head. Or I'm missing an aspect introduced by other parts of the rules.

Unrelated to the Surgery rules; Under Relieve Stress, 7-9, do you remove one stress and choose one, or just choose one? The reason I'm confused is because the first choice says "remove one additional stress" which kinda suggests that you always remove one stress (on a hit).

What noclue and Ebok says, are both good pointers as how to set up the action for Apocalypse World.

Use what they say to create situations that require input. Even if the players do nothing, something will happen. I think its a trap, going into AW expecting the players to push the action. You should push the action forward every time there is a lull. And then when/if the players take over, you can let them determine the course. But if the players don't push the action in the direction they want, then Rolfball the Slaver, who is attacking the hold, will push the action in the direction he wants. And the players will need to do something about that. You shouldn't plan about what they should do. They might defend the hold, they might run away, they might hole up, they might try to negotiate, or go over on Rolfball's side. The game is about watching what the PCs do. But it might be about what they do, in response to your driving forward the action. And if they negate the action, by say, holing up and waiting for Rolfball to move along. Then you push: now the slavers are hacking down the door to the hideout. What do you do? Or: It is three days later. You are out of water and food. You can still hear the slavers partying outside. What do you do?

Consciously work in the fact that moves are made as a result of what the players do. Every time there is an opening for the players to act, ask "What do you do?". If no-one bites or if it makes sense, ask a specific player "What do you do?". If they reply with a move, ask "Yes, but what do you do? How do you <insert move here>?"

I like it. Though I kinda want an extra move to start. Scavenger + 2. When I look at it as a playbook I want to use, I find it hard to pick only one :p Then again that just means that I have room to get cooler with advances. So.. nevermind. One is good, and then I can pick another move with my first advance.

I also kinda want to add a flare gun to the weapon or gear list. I'm thinking (2-harm close reload incendiary). Can't think of a good gear use for it, but it could be handy as gear-as-established.

Other than that, I think you've got a spelling error in "Pick and extra option on a hit with Scavenger". That should be "an extra option" right?