I think it highly likely that without the location they would, as we did, have said dia with little doubt.
This raises an interesting question about identification. It is usually most useful to give an image blind, or even double-blind where you don’t even know who gave it to you. That at least inhibits getting an ID based on where rather than what.
It raises the question, though, that if the circumstantial evidence such as location would aid the process of coming to a more accurate conclusion, would it be helpful to withhold it? I guess it depends on who you are asking. I spent a full day (10.00-20.00) of species analysis with these two experts and they clearly weren’t the sort of people to make off-the-cuff judgements (would have been a shorter day if they had).
Maybe the best approach would be to give an image blind, get an opinion, and then say where and when, and ask if that changes their opinion, not so easy when we communicate by email.
I get a lot of ID requests and I give an opinion based on what I see as the evidence, which I state (maybe a hangover from those teachers who told us to put our workings in the margins). Quite often this arises because someone hopes their image is a rare species (often in the Pyrgus field) and I have to give the honest answer that is often the common(er) species. It is almost Occam’s Razor territory. I do sometimes get the reaction of disappointment and occasionally disbelief.
I concur with your last paragraph, Benjamin, in that it appears that dia and selene can look identical if this image is within range (rather than "a familiar look") for selene. I don’t think we’re missing some ups ID clue. This issue of “within range” comes up so often in the thorny issue of Pyrgus ID. Quite often I give a view that, on the balance of probabilities, an image is 80% likely to be xyz, and I welcome other inputs.
Tristan Lafranchis’ book subtitle was “identifying butterflies is easy”. I’m waiting for his next book “if you don’t understand general relativity, you must be a bit thick”.

Roger