I really don't see what else you could possibly do here besides shove. You've already identified that their range might include a lot of hands (and if you're including any 9x hand, we're really talking a LOT of hands) that you're well ahead of, so with stacks being so short after that big flop bet, it seems like a great spot for you. What other options do you think you have here?

It's not so much that you should worry about folding being exploitable here (it is, but that's not the point because barely anyone in a $2 tournament even knows what 'exploitable' means). It's more that you should worry about it being straightforwardly unprofitable in comparison to other options.

Flatting is a whole lot better than folding. If you're saying flatting < folding then you're effectively saying flatting is -EV, since folding is neutral EV. It's extremely unlikely flatting could ever be -EV unless we a) plan to fold to literally any bet on any further street regardless of the runout, or b) are up against an opponent so tight that we have poor equity versus their flop betting range.

Everything is relative – flatting is bad in relation to shoving because when we shove we get called by so many worse hands (and sometimes villain just folds and we pick up the pot), but flatting can only be worse than folding if it's so bad that our EV is less than zero from this point onwards. If flatting is +1 chip in EV, it's better than folding.

No, I think you're mixing up the idea of 'losing value' with 'being -EV'. EV cannot be 'somewhat negative', EV is a specific number that is either positive, neutral, or negative. Most of the time all we're doing is estimating roughly what this number is and making the play that gives us the highest number, even if we're not specifically trying to calculate it mentally.

In this case, what you've pointed out there is one specific instance in which our opponent check-folds the turn with a hand that would have gotten it all-in on the flop and is worse than ours, which is just a small fraction of all the possibilities that go into consideration when deciding whether the EV of flatting > the EV of shoving or not. In this particular instance, it's disappointing that our opponent might choose to check-fold a hand that could have given us more value, but it's not -EV. It can't possibly be -EV, because in fact on occasions where our opponent folds, our EV is to win the entire pot 100% of the time. We could have won more had we gotten called by a worse hand, but we still won a bunch of chips.

Try not to confuse 'there are some bad things that can happen' with 'this is -EV'. EV already factors in all the bad things that can happen, and weighs them against the good things to create a number that tells us how profitable a decision is. In this hand, both flatting and shoving are +EV, but shoving is significantly more +EV, making it the best play. If you run some calculations with CardRunnersEV, you'll see it's actually reasonably tough to make a -EV play, since usually in any given hand each of the players involved has at least some chance of capturing some of the pot, especially on the flop when both players usually have some equity.