[podiki]

Fine, rather than "underlying theory" use some other words, basically does any new proposal give us predictions, solve some unexplained phenomenon, do something simpler and more compelling (sure, a matter of taste, but so is all of this to some degree), in short, does it tell us something new. Newton certainly did as he gave a law to describe and predict successfully the motions of the planets, for instance. I don't think we are saying anything different really. There wasn't a predicitve framework before (as far as I know), but then there was, so that is certainly progress. But the bar gets ever higher as we know more for what a theory should do. Just writing down an equation that can fit some data and can make predictions is great but is not the end of the story. We could just have some arbitrary functions that fit the data we have and call it a day.

[throwawaymaths]

Yes. Sone researchers went looking for the absence of EFE to disprove MOND and wound up finding it instead. MOND has predictive power. Moreover, we keep finding that lenticular and dense elliptical galaxies have "no dark matter", which is a prediction of MOND.

One of the reasons why things "keep getting added" to MOND is that it's known to be incomplete. The equations to reconcile MOND with relativity are not known and are tougher (but not necessarily impossible) than the Newtonian equations.

So it's expected that things will have to be added. That the equations are tricky for us mere mortal humans should not be held against it; the universe has no requirement that equations must be easy to solve.