I've come to the exact same conclusion as OP. In fact, I actually think small pairs play better with smaller stacks (not larger ones, although obviously we still have to be getting good IO in the pot odds sense) due to smaller stacks can go in more easily postflop with weaker holdings, whereas bigger stacks will only go in with stronger holdings. So there becomes a tipping point where even though the amount we can "win" obviously goes up as stacks get bigger, it is just that far less likely we're actually winning when we get the chips in with a small set (unless playing some really poor players). In other words, I would much rather play 22 with small $150 stacks (50bb in my 1/3 NL game) in a 5way limped pot than I would with $750 stacks going 5way to a $15 raise (even though the IO with regards to what we can win and the SPR are exactly the same in both cases). It's a very weird paradox.
So that combined with the fact that the overwhelming majority of pots are raised at my table (i.e. being able to see a cheap flop to setmine is unlikely unless it has been limped to us in LP), has moved me towards folding small pairs (66-) until mostly LP, and also being very wary of playing them in small juiced pots unless stacks are non-deep.
Obviously the worse your opponents are / the better you are the more you can do whatever you want, but even then their overall profitability (if any) is likely much less than you think (especially OOP) even in the better games.
ETA: If you're getting in 100bb stacks in a limped pot with a set against TP, you're not playing in 2019. Obviously don't waste your time on poker and instead just put huge money on all the upcoming sporting events you know the outcome of.
GbutMikeStarwilldisagreewiththisG