As is often the case, there is more to this than what seems to be the case at first glace.
Apparently the GA aircraft was below the MOA, but more than that, the GA pilot was flying a practice approach and talking to approach, so however you look at it, the GA pilot wasn’t violating any rules or common sense.
As for muscle memory, it’s still quite possible to think you have moved the intended switch/lever/knob/whatever but inadvertently moved one directly adjacent to what you think you’re doing. I know I’ve certainly done this with the light switches in the PA-28 aircraft. And since the control panel in question is apparently not in direct line of sight, I can see that mistake being made while trying to fly formation at the same time while low and slow.
Even with the aircraft still recoverable, my understanding is that the pilot essentially had only a second or two to figure out what was wrong, so I can certainly understand how the pilot might not have realized what actually happened and just assumed the aircraft wasn’t behaving normally and decided it was time to punch out.
I suspect the real issue is probably a variation on the normalization of deviance at that airwing. Normalization of bending some of the training rules here and there to the point where that’s just standard practice and no one really thinks to question it. It’s hard for an individual to go against institutional momentum, and I would think particularly so in a military environment where you generally avoid breaking the chain of command. Things certainly need to change there, but if one looks at all of the events leading up to this mishap, like many, it appears to be the result of multiple chains of events rather than maliciousness or ignorance.