Yeah, basically that happens. I ran into this in school as well, as some professors accepted the current models as reality, which is what made me so sad in gra

Yeah, basically that happens. I ran into this in school as well, as some professors accepted the current models as reality, which is what made me so sad in grad school when I realized how untrue this was. I think of physics as a program. That program is built internally of the various models/equations, and that program takes a set of inputs, which are observables, and a time index. The output is the same set of observables, but with their values at the time index input. That is, physics essentially predicts (forward and backward) from a given set of observations, a new set of observations. So to me, whenever the output observables do not accord with actual measurement, I first assume there must be something in the program that is wrong. Dark Matter is the equivalent of assuming that there must be another set of missing input observables into the physics program. It's possible, but to me the most likely first cause of the inconsistency is the program itself, as that is, as you said, how science works.