In games between two top players. I'm sure it happens a lot in poker but rarely chess. Don't really play chess but perhaps this question is relevant to the controversy.

Your question is hard to follow; are you asking how often a 2400 will find a better move than a 2700 as judged by a computer?

The answer is not very frequently.

Not exactly. I am asking whether positions exist where 2400s IN GENERAL more often (at least in the days before computers were well studied) play what we now are almost sure is the better move than what the very best players do (or did). Maybe like making the two point in backgammon with 64 or perhaps certain limps in poker when first in.

My answer remains the same. I don't think there's a specific type of position that the average 2400 plays better than the average 2700.

Are there exceptions sure - some super positional or tactical 2400 might be better in certain positions than a well rounded 2700 but it's a very individual thing, not a general one.

There probably are a few exceptional positions.

Trainer Jacob Aagaard probably has a couple in his files already. I remember in one of his books he talks about a position that sub 2000 players get correct much more easily than elite players because the best move is quite unintuitive and elite players don't consider it.

Not really talking about a move that no grandmaster would choose thus an amateur would choose it more often. That doesn't correlate to the present controversy.

Given your recent clarification, my answer would be that such a thing would probably only happen regularly in openings where the lower-rated player was more well-versed in. There are many positions where only a few accuracy points separate the top several moves so of course it's easy for a 2400 to randomly pick the better move in close scenarios than 2700s, but that's more luck than skill. For it to regularly happen the lower-rated player would have to know something a super-GM wouldn't, and that's usually only going to happen without going down a particular line of a particular opening. I don't think there are any general positions where lower-rated players would be more accurate.