You make the mistake of thinking that it has to be a MOND theory, I wasn't even thinking of MONDs. There are other models that do not assume that gravity is the only significant force at such levels. Whether these other models can match up to the observations is what we need to investigate. The point here is that none of the current crop of theories matches what we see at all the different levels at which we make observations. This is not a problem per se but an intriguing opportunity to do further investigations into what could be better explanations.
All I am saying is that the assumption of gravity being the only significant force (whatever gravity may be) has only led in the standard models, so far, to a proposal of dark matter for which all experiments designed to find it have failed. We have got to a position where there is a big discrepancy between the evidence that is being collected and what theory is predicting. There are plenty of questions being raised across the board about what the data collected actually means.
I am simply putting forward that there may be a possibility that General Relativity has reached a limit of usefulness and may need to be superseded.
There may be a model that gives an explanation of what we see without resorting to adding entities for which we have not received any confirmation by experiment. What that is, I don't know. I don't get paid to do that kind of research.
If we can't find the experimental evidence to support an idea that theory requires, maybe we need to look elsewhere, including into alternative models and theories.
Unless we are willing to look at different ideas and test them appropriately and rigorously when we come to significant roadblocks, we are then stuck in patterns that aren't helpful.
This is the study of the universe we are looking at, at both the very small and the very large and everything in between.
We make many assumptions in developing our theories and models, but if some of those are not quite the reality then our models can and do give results that seem okay, but in the detail are not.
Mathematical models are always idealised models. They do not take into consideration all that exists in reality. For many purposes, that is acceptable - it becomes a case of close enough is good enough. But, if you take that same model into a physical situation where the idealisations used are not matching reality, then that model will give your incorrect results. For example, in the main you may be able to assume that some feature of reality is of very minor significance, but if you change to a place where that feature is now significant, your mathematics breaks down. You have to change your model or even develop a complete new theory by which you do your investigations.
This kind of thing should be standard practice.
I'll leave you with a final thought. At about the time that Gell-Mann was working on his ideas about quarks, there was another man working on another model to explain what was being seen at the sub-atomic level. Now, this man was not a theoretical scientist but he also came up with an idea of a 1/3 charged particle. His purpose for working on the theory was purely for his own interest (he was dissatisfied with the theory and models around at the time). One he had finished his work, he put it aside and went on with his quite successful life. One of his innate abilities was that he had an eidetic memory and when quizzed could described anything on any page of his 6000+ pages of documents.
I received some of his papers years ago from someone who knew him and had obtained all of those papers. What I found intriguing was that part of the model he developed gave a much better and more easily calculable way of determining the binding energy of any nucleus (from H to Fe at least). He made it quite clear that the model was just that - a model and that one should not expect reality to actually match the model. In another part of the theoretical work he did, he made a slight modification to the electric field equation to include a Lorenz transformation which did away with the need for the strong nuclear force.
Since then, I have come across still another modification of the electric field equation (from a completely different location) which gives rise to a residual force (always attractive) and on the same order of magnitude as gravity.
Now, one of these modifications works on the very small and one works on the very large. Do they have any practical significance? I don't know, but unless there is interest in looking at whether they are capable of explaining what we see, we will never actually know. They may be dead-ends, but they may also open up completely new avenues for investigation.