Another thing going through my mind is that there seems to be a lot of variation in body composition. Fat, flabby person; average person; hard, muscular person, etc. I'm no expert on ballistics gel, in fact I have no experience at all with it, but my first inclination is to think that it, at best, can only simulate one point on the spectrum. I also wonder what materials are closer in variance: ballistics gels or different bodies. I realize the snafu with gels is that if you don't have a standardized product then you have no idea what you're really looking at and you might as well be shooting into a box of leftovers from whatever's on today's specials menu at the local restaurant. However, the human body strikes me as a relatively high-variance thing, and I know first hand that science and engineering models are just that, models, and that every model is only of limited usefulness.
What I'm wondering now is if the G2 has been designed to ace the test, inadvertently or not. You all know how it goes with standardized tests, PFTs, etc. The best way to ace them is to learn how to take them, not necessarily by being amazing at the fundamentals supposedly being tested. If that's what is happening here, then the G2 would be a demonstration in the limitations of current FBI testing protocol. It wouldn't be the first time the FBI came to such a conclusion after some difficult fact-facing in the field: that the test is flawed. The "high-tech" feel of the G2 seems gamey enough to make this possibility not seem farfetched, and we really don't have a ton of hard field data on PT ammo in general, if I understand correctly.
Bear in mind, this is all armchair fun---unqualified pondering. I have no experience with ballistics gel, and I think we are all getting ahead of ourselves. First we should wait to see the data, which should be showing up in the next few months I'd imagine. Beyond that, I agree with the sentiment echoed by a few here that it'd really be wise for anyone thinking about switching to wait on a good deal of hard field data. We already have awesome candidates for carry ammo with battle-proven pedigrees; is anyone really hurting for something better? The FBI has basically made the decision to beta-test this new idea, and we should be extremely grateful for that, because that means we'll get to see a ton of data, data that none of us could generate, for free and without taking the risk. Meanwhile though, it is fun to pass the time playing Sherlock Holmes, and I'm definitely looking forward to seeing some more homebrew tests
I too am curious about the role of the 54226/54227-GOV issue here.