We know gravitationally that dark matter halos aren't cooled on large scales (the size of a galaxy), but just as a galaxy of normal matter has cooled (that's why there are stars and a gas disk) and the normal matter in a cluster isn't collapsed, you could have dark matter cooling on smaller scales.
But getting down to the scale of a snowball cooling without anything *bigger* being affected is difficult. Hard to say it's impossible, but seems very tricky. So in the case that dark matter is a bunch of macro-scale objects, you'd expect to see deviations in larger gravitationally-bound structures.
Unless of course, dark matter just got magically poofed into existence in the form of macroscopic clumps very early on.
I was thinking that it was strange that loosely distributed non-interacting particles would be expected to collapse, but now I'm getting the impression that maybe the opposite is true.
If there's no dark equivalent to EM interaction, and thus no bremsstrahlung, how did dark matter cool enough to collapse as far as it did and play a role in galaxy formation?
Also, a much broader and more basic question that I admit to having never been clear on, what determined the distribution of velocities of the baryons that formed after the big bang? How would thermalization work with particles that don't interact?
Dark matter viralizes through gravitational interactions. That brings the initial overdensities into the roughly spherical halos composes the dark matter around galaxies and galaxy clusters.