Actually, this new theory is credible because there "is" experimental evidence for it (unlike the "dark matter" hypothesis). It explains the motion of stars better.
That’s really not quite true. The observed deviation of galactic rotation speeds from the speeds expected from relativity led to two pretty much independent lines of thought. One of them, exemplified by the theory discussed in this article is that our prior theory of gravity is not adequate. That gives rise to MOND (short for MOdified Newtonian Dynamics). The theory discussed here is a new attempt at a MOND theory. Yes , there is evidence, but it’s not really all that convincing scientifically. These theories have the departure from expected behavior as a starting point and are formulated specifically to account for that deviation.
That is not a bad thing in and of itself, but the real test for a theory like this is applying it to other observations besides the one it was specifically designed to explain. That’s where the problem comes in for MOND theories. If they are correct, they make predictions about things other than galactic rotation speeds. They predict certain primordial abundances of elements and certain patterns of clustering in the cosmic microwave background. The issue is that these predictions fail to match observations. This newest version of MOND might be more successful, but I am inclined to doubt it.
A second line of reasoning exists therefore to try to explain the deviations. In this line, it is believed that our theory of gravity is essentially right, but that the observed rotational variation is caused by unobserved mass. Of course, that is the dark matter idea. It is true that dark matter does not account quite as well for the variations as MOND, but dark matter does a very good job of explaining observed inhomogenities in the CMB and observed primodial elemental abundances.
The jury is still out - neither theory is entirely adequate. Dark matter will be difficult todetect directly, but the idea does have observational consequences, so it cannot be dismissed out of hand. Nobody has yet developed a MOND theory that adequately explains anything other than galactic rotation speeds, but that does not mean that it won’t ever be done. Dark matter is preferred by the physics community right now because of the wider array of observations it accounts for. A MOND theory that does this job would likely find wide acceptance, though, based on the elimination of the need for an unobserved entity. Just don’t make the mistake of assuming MOND is the better theory because dark matter has not been directly observed. There is plenty of observational evidence pointing toward dark matter.