At least from how I understand that, it looks like if I'm going to see a difference in gamma between the Japanese and US releases, it's more than likely going to be that same 10% deal. So it looks like the K-ON!! situation is identical - the number to give to YLevels was 0.900 which made the resultant image match the Japanese release almost exactly. The result of comparing the histogram shift between JP and US screenshots also resulted in a result that'd be close to 0.900, give or take some hundredths or thousandths.
Taking the FMA screenshot used to illustrate it for that release, the math produced something like 0.914, all things considered about the possible inaccuracies of having to find similar structures in a histogram to measure the shift (the result isn't as close as the K-ON!! example either; even using 0.900, the flower things in the background have still have a fairly big difference, even if everything else seems to be a lot closer to the JP screenshot).
Agreed that it's really really stupid to do that nowadays, but at least it seems to be more predictable than what I feared the case might be. I was almost expecting the issue to be worse and the shift to be variable.
It'd still be nice to have a running list of releases that have this issue, unless it's so ridiculously common that it should just be automatically assumed.