One of the unique and testable predictions of MiHsC / quantised inertia is that the dynamics of galaxies should depend on the size of the observable universe. This is because it predicts a cosmic minimum allowed acceleration of 2c^2/Cosmicscale. Why is this? Well, the Unruh waves seen by an object and that (in QI) cause its inertial mass, lengthen as the object's acceleration reduces and you can't have an acceleration that gives you Unruh waves that are too big to resonate in the cosmos. So if you imagine running the cosmos backwards, as the cosmic scale shrinks, more Unruh waves would be disallowed (as in the narrow end of the emdrive), inertial mass goes down, centrifugal forces decrease and so galaxies need faster rotation to be dynamically balanced. Therefore, QI predicts that in the past galaxies should have been forced to spin faster (everything else being equal).
Many people online alerted me to a paper that has just been published in Nature (Genzel et al., 2017) that supports this prediction. The paper looked at six massive galaxies so far away from us that we are looking at them many billions of years ago when the observable universe was much less than its present size, and, sure enough, they spin faster! To compare QI with the data, I have plotted the preliminary graph below.
It shows along the x axis the observed acceleration of these ancient galaxies, determined from Doppler measurements of their stars' orbital speed (a=v^2/r) and along the y axis the minimum acceleration predicted by quantised inertia (a=2c^2/cosmicscale). The QI vs observation comparison for the six galaxies is shown by the black squares and the numbers next to them show the redshift of each galaxy. The redshift (denoted Z) is a measurement of distance. Erwin Hubble found that the further away galaxies are from us, the faster they are receding from us, and so their light is stretched in a Doppler sense and is redshifted. So redshift is proportional to distance. The redshifts of the galaxies in this study ranged from Z=0.854, bottom left in the plot, at which the cosmos was 54% its present size to Z = 2.383, centre right, for which the cosmos was pretty cramped at 30% its present size (the formula for the size of the cosmos at redshift Z is SizeThen=SizeNow/(1+Z).
Quantised inertia predicts clearly that the acceleration increases with redshift, just as observed. The diagonal line shows where the points should lie if agreement was exact. Although the points are slightly above the line this is not a huge worry since the data is so uncertain. The uncertainty in the observed acceleration is probably something like 40% (looking at the scatter plots in Genzel et al. I've assumed a 20% error in the velocities they measured, and a=v^2/r). I have not plotted error bars yet because it'll take time to work out properly what they are. The two highest redshift galaxies are obviously quite aberrant, and this shows that the data is not yet good enough to be conclusive.
So in a preliminary way, and error-bars pending, the graph shows that QI predicts the newly-observed increase in galaxy rotation in the distant past. Given the uncertainties, more data is urgently needed to confirm this. As far as I know, quantised inertia is the only theory that predicted this observed behaviour.
References
Genzel et al., 2017. Nature, 543, 397–401 (16 March 2017) http://www.nature.com/nature/journal/v543/n7645/abs/nature21685.html