Are the LIGO and VIRGO results consistent with the black hole statistics predicted by cosmology?

LIGO and VIRGO detect a very specific subsample of black holes in the universe - those with masses below about $100 M_{odot}$ and which are in close, merging binary systems with other compact objects (another black hole or neutron star). There is a strong bias towards detecting the most massive of these systems because they emit more powerful gravitational waves (GWs) when they merge.

In particular, they do not detect: isolated black holes, since they produce no GWs; very massive black holes, even if in binary systems or accreting other stars, since the frequency of the GWs produced is below the detection thresholds of the instruments; or black holes in wider binary systems that will not merge on timescales shorter than the age of the universe.

Because of these biases it is unlikely that LIGO/VIRGO will be able to make any measurement that will assess the total numbers of black holes in the universe. However, that is not to say that they can't test the details of stellar evolutionary theory. For example the mass distribution of the black holes detected by LIGO/VIRGO should be a powerful test of ideas for how very massive stars end their lives - in particular, there may be a high-mass cut-off at around $50M_{odot}$ arising from the pair instability process (e.g. Belczynski et al. 2016; Farmer et al. 2019) and LIGO/VIRGO are providing significant evidence on the relative scarceness of black holes with masses $<5M_{odot}$. The observations also play a crucial role in understanding how these binary black holes form in the first place, either from massive binary stars or perhaps by capture in dense environments.

So in that sense, the LIGO/VIRGO observations are leading the theory towards making better predictions about the black hole population because the theory had many uncertainties and assumptions built into it.

In terms of, have enough GW events been detected? The answer is surely no. From my perspective, it seems to be the boundaries of the distributions that are of most interest, or the rare events like the neutron-star merger that produced a simultaneous kilonova. Almost by definition this means that you have to detect very large numbers of "median events" in order to properly probe the limits of what is possible and seek out the most extraordinary events which may give the most interesting information.