Let me try to back up another step. When Einstein (and others) laid the groundwork for the relativistic model of gravitation, the nature of our visible universe was very poorly understood, and neighboring galaxies were still called "nebulae" and their natures were unknown. Newtonian gravitation explained the behavior of solar-system bodies pretty darned well, and the relativistic interpretation was marginally better.
Here we are ~100 years on, and we have lots of wonderful observations. Vera Rubin discovered from spectroscopic studies that the rotation curves of spiral galaxies are much flatter than can be explained by GR's gravitational model. The MOND folks introduced modifications to the Newtonian gravitational model that not only fit the flat rotation curves very well, but also predicted the behavior of low-surface brightness galaxies over a decade before they were actually observed. This does not prove that MOND is correct, just that is predictive and therefor somewhat useful.
On larger (cluster) scales we have observed that the gravitational binding forces are larger than can be expected from a GR model using visible matter. Zwickey knew this a long time back, so this is not a new problem. Lensing effects are also much stronger than we might expect by considering only the visible matter.
We also have the Pioneer anomalies, showing unmodeled Sun-ward accelerations in spin stabilized spacecraft . Is it possible that neither the Newtonian nor relativistic models for gravitation are accurate on some scales? I think that we should be open to the possibility that these gravitational models might have a limited scale of applicability. Invoking the existence of exotic undetectable matter that only interacts gravitationally is a kludge. Even if it were a viable fix, it is advisable to look at what the MOND folks have done and ask what we could do to modify current gravitational theory to more closely accord with observations.
I do not have a blind allegiance to Einstein or his work, despite what you might think. I am trying to point out that what he accomplished has to be seen in the context of scientific knowledge at that time, and inventing new entities (DM for one) to try to keep his relativistic model of gravitation predictive is short-sighted and likely counter-productive. Reading Einstein's tribute to Ernst Mach, I get the feeling that he would be very uncomfortable with this situation.
"How does it happen that a properly endowed natural scientist comes to concern himself with epistemology? Is there no more valuable work in his specialty? I hear many of my colleagues saying, and I sense it from many more, that they feel this way. I cannot share this sentiment. ... Concepts that have proven useful in ordering things easily achieve such an authority over us that we forget their earthly origins and accept them as unalterable givens. Thus they come to be stamped as 'necessities of thought,' 'a priori givens,' etc. The path of scientific advance is often made impassable for a long time through such errors. For that reason, it is by no means an idle game if we become practiced in analyzing the long common place concepts and exhibiting those circumstances upon which their justification and usefulness depend, how they have grown up, individually, out of the givens of experience. By this means, their all-too-great authority will be broken."