This^^^^^^^A couple of comments.
Let's assuming that the AMP annealed cases are better in this situation, the question is was your gas annealed cases not annealed properly or annealed properly but the annealer is not capable of annealing as evenly as the AMP?
The other thing is coming from working with human clinical trials of drugs, testing of this sort is always fraught with the "placebo effect" since the shooter knows which case were which. As a suggestion, if you really want a reasonably un-bias study, one you will need a larger sample size, more than one group (since as you know, even you shooting the cases done with one machine don't always shoot the same group size over and over again) and finally do the study in a semi-double blind fashion, which is to have another shooter hand you the round with only him knowing which group is which but not the shooter.
we all want better, but you have to do solid testing with a lot of this stuff to truly know if its better. a double blind test with large sample sizes......and no, a couple 5 shot groups wont tell you anything imho