> where we would put matter if we could to make our theory of gravity make sense
Dark matter behaves in a fundamentally different way from baryonic matter. We can constrain the total amount of matter in the universe (both dark and baryonic) from the observed abundances of baryogenesis. But dark matter has a different effect on the relative amplitudes of peaks in the CMB.
As far as I can tell, MOND has never really had any success outside of modeling galaxy rotation curves.
The skepticism I've seen towards dark matter vs. MOND has always been strange to me. Dark matter doesn't really require much in the way of new physics --- there's just a new particle to add to the standard model. But most MOND theories violate Lorentz invariance which is a vastly more radical departure from standard physics. (And in my mind, the more sophisticated MOND theories that maintain Lorentz invariance like TeVeS are really a theory of dark matter dressed up in the language of MOND.)
These successful predictions are all generally variants on modeling galactic dynamics, though. The trouble is that galaxies and galaxy clusters are very messy places, so it's hard to make sure you've incorporated all the relevant physics.
By contrast something like baryon acoustic oscillations are very simple to model, so you can be quite confident that you've incorporated all the relevant processes. And in that regime LCDM performs beautifully and MOND completely fails. So it's reasonable to suspect that in more complicated environments the problem is that we're not modeling the systems correctly rather than that there's new physics going on.
MOND doesn't cover the existence of CBM, distribution of galaxies, non-metallic abundance - things all covered by LCDM.
What MOND has going for it is that galactic rotation curves are readily consumed by popsci readers and the story of the "little guy" vs the scientific establishment is an easily available frame story popsci authors can sell clicks for.
The proportion of lay people who think MOND could be true greatly outnumbers the proportion of MOND researchers and doesn't reflect the veracity of the theory.
MOND is not a cosmological theory unlike LCDM, and it isn't relativistic. So we should not expect it to cover the range of things that LCDM tries to.
It's just a tweak to Newtonian gravity, which surprisingly matches observation very well, and has accurately predicted quite a few things in the regime it operates in, before they were observed.
The fact it works so well in the areas it does apply to is the reason that science hasn't given up on it yet (regardless of what pop science or lay people think).
Dark matter behaves in a fundamentally different way from baryonic matter. We can constrain the total amount of matter in the universe (both dark and baryonic) from the observed abundances of baryogenesis. But dark matter has a different effect on the relative amplitudes of peaks in the CMB.
As far as I can tell, MOND has never really had any success outside of modeling galaxy rotation curves.
The skepticism I've seen towards dark matter vs. MOND has always been strange to me. Dark matter doesn't really require much in the way of new physics --- there's just a new particle to add to the standard model. But most MOND theories violate Lorentz invariance which is a vastly more radical departure from standard physics. (And in my mind, the more sophisticated MOND theories that maintain Lorentz invariance like TeVeS are really a theory of dark matter dressed up in the language of MOND.)