[dnautics]

Yes but we have independent observation of that baryonic matter. Suppose we hadn't discovered the theory of relativity. The rules for gravity hold everywhere except for the orbit of Mercury. For several decades it is reasonable to guess there is "mystery meat" planet somewhere that we can't see. Possibly even a hamburger sized black hole orbiting the sun. But to think that the field equations for gravity are wrong is nonsense. Fast forward a few decades and Eventually we can't find this mystery planet and it's decided that an interpretation that the gravity field equations are wrong is more sensible.

[matthewdgreen]

I’m not objecting to the idea of alternative theories. I’m objecting to the parent commenter’s idea that “explains more galaxies with a single parameter” should be an argument in favor of it. By that metric your Mercury example is a great illustration: oddities in several of the outer planets’ orbits were caused by unknown planets, but only Mercury’s unusual orbit requires new theories of gravity. So clearly the Newtonian theory explains more instances with a single parameter, and is thus better. (Obviously in the real world this is not how theories should be evaluated.)

The appropriate MOND/DM analogy here would be to propose general relativity as an explanation for Mercury but then find another planet(oid) in the system where GR suggests an effect, but to notice that the effect is measurably not present. You can’t just say “got most of them right” when you propose such a theory. You need to revise your theory until it explains the observations.

[daxfohl]

That's quite a good point. I hadn't considered that perspective. I'll have to think about that one.