The Lambda-CDM cosmological model is succesful at reproducing various independent sets of observations concerning the large-scale Universe. This model is however currently, and actually in principle, unable to predict the gravitational field of a galaxy from it observed baryons alone. Indeed the gravitational field should depend on the relative contribution of the particle dark matter distribution to the baryonic one, itself depending on the individual assembly history and environment of the galaxy, including a lot of complex feedback mechanisms. However, for the last thirty years, Milgrom’s formula, at the heart of the MOND paradigm, has been consistently succesful at predicting rotation curves from baryons alone, and has been resilient to all sorts of observational tests on galaxy scales. We show that the few individual galaxy rotation curves that have been claimed to be highly problematic for the predictions of Milgrom’s formula, such as Holmberg II or NGC 3109, are actually false alarms. We argue that the fact that it is actually possible to predict the gravitational field of galaxies from baryons alone presents a challenge to the current Lambda-CDM model, and may indicate a breakdown of our understanding of gravitation and dynamics, and/or that the actual lagrangian of the dark sector is very different and richer than currently assumed. On the other hand, it is obvious that any alternative must also, in fine, reproduce the successes of the Lambda-CDM model on large scales, where this model is so well-tested that it presents by itself a challenge to any such alternative.
Date added: Wed, 16 Oct 13