theory makes sense, but i think its a bit like its been concluded upon in a vacuum. How often do you expect to see appropriate stack sizes for this to make sense? If both stacks are short this will rapidly result in either the end of the tourney or one very small vs one average stack.

The push/fold situation occurs in all tourneys as blinds increase and players are forced to bet or bleed their whole stack in blinds. Which often leads to an ever increasing % of hands that are pushed. If there are big stacks around you start seeing players call with worse hands just because its so cheap to bust the villain. I think for your theory to hold true you need to have a short stacked cash game, like a max 30bb buyin.. there's certainly not going to be alot of those around.

Optimal strategy vs shortstack push folder is likely to be deepstack, fold/call. And if the deepstack makes that particular adjustment then it would be in the shortstacks best interest to not push, but rather just make a standard raise for 3-4bb. No point risking the whole stack if your just going to face a hand that either folds or is willing to go all in. Thus begins the never ending spiral of adjustments in heads up poker that makes it a game where no optimal strategy exists based on cards. Only the idea that you should remain aggressive and correctly adjust to your opponents tendencies, whatever stack sizes are involved.