The best-well known effect is the "missing mass" in rotating galaxies. Well recently we found galaxies that look like they have no dark matter in them - i.e. they behave as you would expect from classical gravity/centripetal force arguments [1]. So while you can tweak General Relativity to match some of the effects in typical galaxies, you would then expect the results to hold in all galaxies. The "outliers" are easily explained by Dark Matter - they just don't have it for whatever reason.

The simulations of Universe formation, sure have their limitations, but look much worse without DM.

Galaxy clusters look like they have unexplained mass in them. Crucially, this isn't only from observing their rotational speeds (so like galaxy rotations) but also from looking at the strength of gravitational lensing they induce. This is as compared to fluid-dynamics-y calculations of gas temperature, pressure etc.

Bullet cluster is another interesting case [2]. It is the result of a collision of two galaxies. In such cases, you expect (a) the stars to "fly through" each other, as they are basically point masses and only interact gravitationally, but (b) the gasses and dust to collide. In this case, most of the mass is in fact gas and dust, so it should be distributed in the middle of the cluster (having just had an inelastic collision). But in fact observed mass is where the stars are, further out of what looks like center of mass.

[1] https://www.nature.com/articles/d41586-022-01410-x [2] https://en.wikipedia.org/wiki/Bullet_Cluster

But GR is complex enough to allow very different behavior depending on the initial conditions of galaxy formations. So it would not be surprising to see different scale of deviations from Newtonian mechanics depending on Galaxy history or age.

[1] - https://arxiv.org/abs/1810.04445