SPACE.com posed some complicated questions to the University of Cambridge's Neil Trentham, who responded with in-depth answers:
SPACE.com: Why might a galaxy of dark matter not have stars? If "normal" galaxies are loaded with dark matter, as most scientists believe, then would you presume that there is some relationship between the presence of stars alongside dark matter?
Trentham: There is a relationship but it is indirect. Galaxies started out in the early universe as small dark matter lumps and then grew into larger galaxies by gravitationally pulling in more dark matter and diffuse gas (not stars) from intergalactic space. If all this happened in the very early universe, then the dark matter halos can pull in quite a lot of gas, since the gas is relatively cold. A lot of cold gas is quite easy to transform into stars, by further collapse of material within the halos.
The end result is a normal galaxy.
If all this happened at later times (say, in the last 5 billion years), the little halos still grow by attracting dark matter, but they cannot pull in much gas. The reason for this is that the gas between the galaxies has now been sitting there for a long time and has been heated up by all the light from the large number of galaxies that have formed over the age of the universe. The hot gas is difficult for the halos to pull in, since the gas atoms are moving so fast.
The end result is a galaxy with lots of dark matter but no stars -- a dark galaxy.
In summary, the important issue as to whether or not a galaxy is dark is likely to be the ability of the dark matter to attract gas from the space between galaxies which later turns into stars.
SPACE.com: What is the likelihood that dark matter is merely the combination of brown dwarfs, black holes, and other dead stars, as opposed to some unknown fundamental particle?
Trentham: I would say quite unlikely, but the reasons are a bit indirect. There are two:
1) The total amount of dark matter in the universe is determined by counting the total number of galaxies and attributing a mass of dark matter to each. These masses are determined by dynamical measurements. The total amount of [normal] baryonic matter (of which stars, brown dwarfs and the various kinds of stellar remnants are made) is determined from Big Bang theory, which is very well tested and in my mind is quite robust. The measurements relevant to the baryonic density come from helium and deuterium abundances, which are reasonably well determined.
The end result is the total density [from the first method] is several times the baryonic density [from the second method]. So I doubt all the dark matter is baryonic. There are of course many caveats, but the vast majority of evidence supports the above view.
2) If all the dark matter is baryonic, then making galaxies at all through gravitational growth of halos is difficult. So is predicting the large-scale distribution of galaxies. Both of these things are predicted from theories where the dark matter is non-baryonic, however. This result comes from the output of many computer simulations.
In summary, I'd say that all the evidence in conjunction is pretty conclusive, though each detailed observation or theoretical result is open to question.