1) It sure looks like the .32 S&W Long, especially with, say the Buffalo Bore load (100 grain .32 caliber WC at maybe 825 ft/s) is more potent than one-half a duplex load (say 100 grain .38 WC at 600 ft/s). The .32 would have 24" of BG penetration; the .38 less than half that. While in terms of wound trauma mass the two are more or less equal at about 19 grams each.
But there are two wadcutters in the duplex load. Of course, that doesn't improve the penetration problem (more on that below) but it does double the wound trauma mass to close to 40 grams. More importantly, if it can be made to work, the duplex load might, as MacPherson wrote "greatly increase the chances that at least one bullet will penetrate a vital structure." (
Bullet Penetration, page 52).
2) Yes. The CG vs validated 10% ordinance gel is one problem and the inadequate penetration of 100 grain .38 WC at 600 ft/s is another. I think there's a very simple workaround for the first problem. But the second problem, is simply an accident of physics, and that too can be easily solved.
I'm not married to 100 grain WC at 600 ft/s. What if we just move to a 110 grain WC? Then the predicted BG penetration jumps to just over 12" and satisfies the FBI minimum requirement. Sure, not by much. But why not pump the velocity up to say, 650 ft/s? That gives just over 13" of BG penetration. Still not enough? Then replace the front wadcutter with a 95 grain truncated cone. Then the front bullet has over 15" of penetration. That should be enough.
I know there are folks who question the "paper" predictions and apparently don't trust the penetration models of Mssers. MacPherson and Schwartz. But I do.
3) Rightly or wrongly, I haven't worried too much about bone. I figured hard-cast, sharp-edged wadcutters should resist glancing better than a FMJ and deformation better than a soft factory target wadcutter. And I think the FBI 12" penetration requirement already takes some of the bone problem into account.
I can't answer the dispersion question with any authority as I'm postponing such question while I work on the penetration problem. But I have shot plenty of targets over the chronograph at five yards which can at least give one data point.
This was was five shots from a "hybrid" load -- a 95 grain TC in front and a 105 grain WC in back at 700 ft/s. Because the WC and TC cut different holes, we can see that even at 5 yards there is some obvious vertical dispersion between the front and back bullets. I assume that the back bullet leaves the barrel an instant later and is more effected by recoil. But I'm not sure of that.
At any rate, I've never seen any tendency for the bullets to "zing off in different directions."
The notes on the target show the predicted penetrations (TC=16.5", WC=13.1") and under the "NRA" the average measured penetrations in CB gel: TC=19.5", WC=14.5". Those last numbers, now highly suspect, led me to think I was seeing a "drafting" effect. I hope to soon have that sorted out.