If a hand isn't a profitable 3b, then you shouldn't 3b it.

0 EV is a thing..

The onus is generally on the person asserting the absurdity to show cause. Please prove to me that 3b 93s is >= 0 ev.

Ok let's say you have a 100bb open shove range of 100% of AA combos and 1% of 93s combos. Hopefully you can agree that people should be folding everything but AA vs this range. Therefore when you do shove 93s it will not be -EV. The same logic from this extreme example applies in a 3bet situation.

First, your example is only true vs an opponent exploiting your shove range. Vs someone playing reasonably jamming 93s is of course a huge mistake. Second, I don't think an example where you purposely play badly to make a particular hand +EV is very useful in proving your case.

I do think the reasoning you give ("add them in occasionally to balance") is the reasoning a lot of people use, but that does not make it true!

its nice to add hands that are slightly too weak to call i think. 93s is a little much, but a hand like 95s is decent if you are generally defending 96s+ (lets say to a 2.5x-3x rather then 2.2x). 95s will have close to zero EV as a call and it can be used more effetely in a 3b range to enable your value hands to get called more pre flop. it keeps calling range strong and has decent barrel potential and board coverage features

I tend to use stronger suited connectors, hands like T9s-T7s, 98s-96s, 87s-85s, etc plus some weak Axs/Kxs. All of those hands I do some mix of calling/3betting. Generally speaking I don't like the term bluff preflop, even the weakest hands have pretty reasonable equity. But I think using the stronger semi-bluffing hands makes sense vs someone defending aggressively/appropriately against 3bets.

Good question that I don't have a great answer to. Some math you can skip:

Let's say he 4bets 15% of the time, folds f%, calls 1-f%, our EV when called is x.
EV = 0.15 * (-19) + (1-f) * x + f * 9
For EV = 0
0 = -2.85 + (1-f) * x + 9f
x = (2.85-9f)/(1-f)

When called pot will be $42 (less rake), we risked $19.
If he folds to 3bet 60%, then breakeven point is lose $6.38 when called. We need to win ~30% of the pot on average.
If he folds to 3bet 50%, BE point is lose $3.30 when called. We need to win ~37.5% of the pot on average.

Side note: winning 37.5% of the pot is NOT the same as winning the pot 37.5% of the time.

Rough estimates, but I think if he's folding 60% you can 3bet profitably, if he's folding 50% probably not.

Let's say he 4bets 15% of the time, folds f%, calls 1-f%, our EV when
called is x. EV = 0.15 * (-19) + (1-f) * x + f * 9 For EV = 0 0 =
-2.85 + (1-f) * x + 9f x = (2.85-9f)/(1-f)

I think you made a mistake here, he calls 1-.15-f% amirite? so you'd end up with x=(2.85-9f)/(0.85-f)

60% folds means you can lose $10.2 when called or win 21% of the pot.
50% folds means you can lose $4.7 when called or win 34% of the pot.

Rake at lower stakes would increase the win%'s required by 1-1.5 %-p

Winning 30% is enough with 54% fold vs 3b%.

So maybe 50% folds is actually enough or thereabouts?

Yep, good catch. Hm I'm a bit skeptical that 50% is enough from a gut feeling perspective, but perhaps doesn't need to be much more. Also of note is that we've ignored bb cold 4betting, so we get 4bet an extra couple %.

Not sure what you're asking, I have no program that shows me GTO plays. This might be the video I looked at some of Poker Snowie's preflop strategy (?) but Snowie does not play GTO.

Don't think I'm really qualified to say either way. I did buy the pro version and have not gotten my money's worth, but that's largely due to me just not spending nearly as much time with it as I'd planned to.