10-12" definitely indicates overexpansion or fragmentation. The optimal penetration range against humans of average size is 12" or greater, with an average of 13" or more being ideal. The optimum is somewhere around 13-14". An 85 gr bullet at 800 fps can only expand to a hair under .39" and still penetrate 13". That's only 1.25X expansion. Very, very few bullets will actually expand a small enough amount, reliably, to get decent penetration. The Hornady 90 gr .380 XTP comes close, with consistent 1.3x expansion, but that still limits penetration to about 11".
My calculations were based on Federal ammo, which hits 1100 fps through a 4.5" vented barrel. Assuming a chamber pressure of 14,000 PSI to 17,000 PSI (instead of just 14,000, like I asumed before), the range spreads out to 800-880 or so. Let's see, Georgia arms says 100 grains at 1,100 fps, no barrel length given. If it's 4.5" again, I get 800 to 860 fps, actually. A little slower.
Well, let's see what .38 SPL does through this calculator. 16,500 PSI, 158 gr, 755 fps, 4" vented barrel...
4" - 755
3" - 716
2.5" - 690
2" - 659
1 7/8" - 649
100 fps slower, so that's close to the "rule of thumb."
For .38 +P... 18,000 PSI, 158 gr, 890 fps, 4" vented barrel...
4" - 890
3" - 835
2.5" - 799
2" - 756
1 7/8" - 744
Though most manufacturers don't push the upper limits of the .38's pressure like that (both PSI figures are 500 PSI below the SAAMI maximum). Some .38 +P handloads top out at 17,000 PSI. For safety, I guess. That's also assuming that the vents don't shorten the effective barrel length to 3.5" or 3.25". Why do manufacturers always test .38s through vented barrels? Anyway, I'll run some more numbers with actual PSI figures from a reloading manual later.
Edit: Oh, wait, crap. I totally forgot that SAAMI chamber pressure standards (and reported pressures in load books) are the average pressure. This crazy calculator uses the peak pressure for its input. I'll need to totally recalculate all those velocity figures. They'll probably be much higher.