More accurate observations of globular clusters have turned up low luminosity stars, with masses of about 0.18 times that of the sun, that account for a large proportion of the globular cluster's previously estimated dark matter which was based on brighter observed stars and lensing observations for the entire cluster.
These clusters are considerably more rich in dark matter than individual galaxies and the phenomenological predictions of modified gravity theories derived from an early version called MOND consistently underestimate the amount of dark matter in these clusters. But, this new result finding that there is considerable luminous ordinary matter in these clusters that had not previously been seen by astronomers because their instruments weren't powerful enough to see them suggests that MOND's shortcoming in its estimates of the magnitude of effects due to something other than Newtonian gravity acting on observable luminous matter may be much smaller than previously believed.
This result should be considered together observations in late 2010 that revealed that the amount of low luminosity ordinary matter in elliptical galaxies had been grossly underestimated. The more accurate ellipical galaxy census suggested that the true amount of dark matter in the universe due to that revision in the estimate amount of normal matter in ellipical galaxies alone was closer to 50% of all ordinary and dark matter combined, rather than the frequently quoted 80% figure.
Other recent theoretical studies have shown that some portion of the effects attributed to dark matter in spinning galaxies is actually attributable to general relativistic corrections to models that estimate the effects of gravity with a Newtonian approximation, although different theorists have reached dramatically different estimates of the magnitude of these effects by using different methods to simplify the description of a spinning galaxy to which the equations of general relativity are applied.
This new result on globular clusters, combined with the prior work on dim matter in ellipical galaxies and general relativistic effects, suggests that the actual percentage of matter in the universe which is dark matter may be considerably less than 50%. Dark matter may actually end up being one third or less of all of the matter in the universe.
If one takes the position that a cosmological constant is a perfectly acceptable and respectable alternative to a hypothesis that 80% of the universe is made out of "dark energy" observed in no other way, and that it represents a property of space-time itself, and that the actual proportion of matter which is dark is much smaller than previous estimates, then dark matter candidates like neutrinos (perhaps in condensate form) that don't require the discovery of new fundamental particles, begins to look more plausible.