[isthatafact]

What you are calling a model is simply a way of quantifying a measurement of a galaxy to describe (or model) its mass distribution. There is no physics involved.

> "DM theory needs to make predictions we can then test"

That is what scientists do: Start from the hypothesis of a dark matter dominated universe, from the beginning (or soon after the big bang), then turn on time and physics (gravity plus gas physics in a computer simulation), and galaxies form as gravity causes matter to clump together. The properties of those theoretical galaxies are testable predictions.

[throwawaymaths]

You're confusing the broad scale cosmological predictions of DM with explaining rotational parameters of individual galaxies.

Even so, there are predictions low-parameter DM models seemingly can't make, like: what percentage of galaxies have zero dark matter?

[isthatafact]

If you were to create any model of the universe, you would require it to have the correct (observed) percentage of galaxies without dark matter, right?

You assert that it is impossible for a model of a dark matter dominated universe to "predict" galaxies without dark matter, but that is exactly what people are looking at here in a large scale cosmological simulation: https://arxiv.org/abs/2202.05836