[ldunn]

This tells you that the chance of detecting dark matter is higher than it would be if it were only, say, 0.1% of the energy content of the Universe. Which is true, but so what?

Maybe you mean that this is a sense in which how hard it is to detect has something to do with how much of it there is, which is fine, but the original question is suggesting some much more constraining relationship, where the fact that dark matter is hard to detect and the fact that there is a lot of it poses some kind of apparent contradiction, or at least a puzzle. I don't know of a reason to think that any such contradiction exists.

[naasking]

No detections for abundant stuff over long time frames is not a logical contradiction, but it certainly decreases the plausibility of such a theory relative to alternatives, particularly since we're not expecting any such particle for any other reason (we've tested broad energy ranges). Literally the only evidence we think it exists is incompatibility of our gravitational theory (GR) with observations in a regime where this theory is not as well tested. We already know GR is wrong for other reasons, so the real question is, on balance, which is more plausible? Maybe in the 80s the balance of probability swung one way, but given the unexpectedly successful predictions MOND has made, the repeated unexpected prediction failures of DM, the new energy regime we've probed in the LHC showing nothing, and the repeated, failed DM detection experiments, it's very reasonable that people might conclude that the balance has shifted the other way.

[ldunn]

Er, again, fine, but this has nothing to do with the question of "how can something so hard to detect be so abundant". The question isn't "should the non-detection of dark matter decrease my credence in it".

It's of course true that the continued non-detection restricts the available parameter space. But there is no physical principle that says that if something is as abundant as dark matter it really ought to have been detected by now - it's not as though if something is as abundant as dark matter, then it really needs to have some minimal coupling to baryonic matter that the existing experiments are now ruling out. Dark matter can just be really very hard to detect, there's no issue of how that "can be". Things that can change the rotation of galaxies are not obliged to be detectable by 2025.

If your weighting of the relative probabilities is such that you feel you should be going to bat for MOND at this point, that's your prerogative. But it's not related to the question that was asked.