I don't disagree with much of anything you're saying (not much anyway--chimp society is complex enough that you can look at different parts of it and argue about it almost as endlessly as human society); I was just focusing on hierarchy depth and stability as the measure of the degree of "dominance hierarchy". The key feature not being dominance (because when there isn't any established structure, competition is still solved by a type of dominance--physical aggression if threats don't work), but rather hierarchy.

However, I'm not sure you've pinned the problem with males-in-unstable-conditions on the right thing. Males could just a be a problem when there's instability, period. Do we have enough examples of instability without a dominance hierarchy default so that we have good evidence of what default male behavior is there? Human behavioral plasticity is so large that I wouldn't be surprised if we could find a single instance of almost everything, but if you go back to historical evolutionary pressures, it seems like "selfish violence when you can get away with it" would potentially be a selected-for trait when there was any manner of societal breakdown. Of course, that doesn't mean that humans are actually that way; it just motivates asking the question.

(I can't tell if you meant to say that men rape during war in order to "return to a sense of stability"...that's what it reads like to me, but it makes so little sense that I think I've misunderstood and you meant something else.)