General relativity is only a good approximation for extremely limited scales, on the order of solar systems.
It's applicable area of operation is .1% of the universe.
This is easily demonstrated. The theory by itself is 99.9% accurate within the solar system without adding any other theory to it. But the second one attempts to apply what was just shown to be 99.9% accurate in describing planetary systems (non-ionized matter) to the rest of the universe (ionized matter) one suddenly has to add 96% ad-hoc theory to make what was just tested to a 99.9% accuracy without it to make it even fit a semblance of reality.
So the question is really, are we using the wrong physics for the wrong state of matter to begin with in our attempt to describe the rest of the universe, and so end up needing that 96% ad-hoc theory leading to singularities?