Does anyone know a ballpark aggregate frequency for busting your stack in various situations?

Why do you think this would be useful to know? In a MTT you should be focusing on whether a play has a positive $Ev (which in the earliest stages of the tournament will be very close to chipEv). If you are up against a stack (or multiple stacks) that covers you that sometimes mean you will bust quite often. If a play would result in you bust too often it will have a negative $Ev.

When considering SPR the more useful approach to me seems to be trying to learn what kind of hand it normally will take to be willing to play for stacks.

I think about it in spins vs recs lets say I have 33 in the BB 25 deep, its a mix between flatting and jamming. Lets just make up some #s and say I get called 50% of the time with 50% equity If I jam so I bust 25% of the time where as flatting even though its theoretically the same ev I have virtually no chance of busting.

I've thought about the same spots in mtts as well where gto mixes 3betting and flatting and I imagine 3betting probably has an increased chance of busting which is bad but..... flatting also leads to multi way pots which can be almost impossible to win so who even knows if it "reduces variance".

When 25bb deep in BB with 33 solvers often want to shove 33 100% of the time. Small pairs realize their equity best either with deep stacks when the implied odds makes set mining profitable or by putting the opponent allin with shallower stacks.

Gtowiz has it as a 66, 2.3, 31.7, jam/3b nai/flat mix. That wasn't the point. Even if all 3 options have the same theoretical EV 1 option does lead to more bust outs which might be bad just like it might be bad in MTTS so it might be useful information to know.

I don't personally have the database, but it seems like the "easy" way would be to look at actual statistics using MDA?