That is a kick-ass hack.

And yeah, the more dice you roll, the more results will clump toward the average. So as long as the average doesn't veer too far away from 7-9, you're good. So that 2d6 + 1d4 roll (which averages 9.5, as compared to the 2d6+3 and 1d6+1d12 rolls that I broke down above, both of which average 10) looks like this:

14% chance of 6- (miss)
36% chance of 7-9 (weak hit)
50% chance of 10+ (strong hit)

Whereas 3d6 or 1d4+1d6+1d8 (which both average 10.5) look like this:

10% chance of 6- (miss)
28% chance of 7-9 (weak hit)
62% chance of 10+ (strong hit)

(The probabilities aren't exactly the same.  But they're within a percent of each other)

It's not about the average, though.  Let's look at a 2d6+1 vs 1d6+1d8:

with 2d6:

28% chance of 6- (miss)
44% chance of 7-9 (weak hit)
28% chance of 10+ (strong hit)

vs 1d8+1d6

31% chance of 6- (miss)
38% chance of 7-9 (weak hit)
31% chance of 10+ (strong hit)

As you can see, what this does is increase the chance of a miss or solid hit, but  at the expense of a weak hit.  The effect is even more pronounced at the extremes:

2d6 + 3
8% chance of 6- (miss)
33% chance of 7-9 (weak hit)
58% chance of 10+ (strong hit)

1d6+1d12
21% chance of 6- (miss)
25% chance of 7-9 (weak hit)
54% chance of 10+ (strong hit)

2d6-2
72% chance of 6- (miss)
25% chance of 7-9 (weak hit)
3% chance of 10+ (strong hit)

1d6+1d2
75% chance of 6- (miss)
25% chance of 7-9 (weak hit)
0% chance of 10+ (strong hit)

I think Vx looked at probabilities pretty carefully when he designed the system.  The weak hits are usually the most interesting results, giving us both player success and interesting consequences, and the change you propose makes them more rare.  It also makes misses more common in general.  If doing either of these things was part of your goal, then more power to you.  But if you wanted to simply provide a more interesting dice mechanic while maintaining similar results, then your proposal doesn't do that.