It's amazing that every dark matter discussion goes the same way, with the same amnalogies to aether or epicycles, completely unlinked to the actual evidence.

It's amazing that every dark matter discussion goes the same way, with the same amnalogies to aether or epicycles, completely unlinked to the actual evidence. (Unlike the blog post, where Hossenfelder suggests a specific methodological problem, and that GR may in fact be completely correct. This is completely unlike anything people have suggested here in the comments.) Dark energy is less well established, but based on the evidence we have, Occam's razor points strongly towards dark matter. Dark matter is a specific theory that makes predictions, and as far as we can tell these predictions are borne out. There was a competing hypothesis -- that GR itself is wrong and needs to be corrected. These were in competition, but the evidence has decisively broken in favor of dark matter. If it was just a question of fixing the equations, then you should be able to predict the apparent pattern of dark matter from the distribution of ordinary matter. For a while it looked like maybe you could, but more recent evidence has shown that you can't. (This is the reason why the Bullet Cluster comes up as an example so much -- the Bullet Cluster looks like what you would predict if galaxies have dark matter halos and two galaxies collided.) People also misunderstand the incentives here. If you could write down a modified theory that could explain all of the evidence, you would be the greatest physicist since Einstein, maybe the greatest since Newton. You would achieve immortal glory as one of the central figures of the 21st century. So people try to write down that theory, but it turns out to be really hard to do.