Earlier quoted context omitted.
This carries the implicit assumption that all forces fall off monotonically with distance. Maybe gravity doesn't? Maybe it has a curve similar to a meteor impact crater: raised in the middle, a deep valley, and raised at the outer edge? Of course, with more evidence, such as the observations of the Bullet cluster, these simple explanations fall apart. But at any rate, it's not enough to take one observation and assum…
Gravity behaving like that would be a lot weirder than there simply being some types of mass we aren't able to detect yet. Physics has had a long history of predicting particles that hadn't been detected yet, and then actually going out and finding them. See neutrinos, the Higgs boson, etc. General relativity being wrong would be waaaay weirder.
If I'm not wrong, Newton's laws can still be held to be true in some cases even inside GR?
Some where in the past, some one could come up with a new model which could show us new truths, and yet could contain GR to be true in some cases.
In GEB, Hofstadter show us part of the ways so see intelligence emerging from noise and garbage, is to understand when you design a system, you start with axioms, truths and rules that are true in a system, could be total garbage in another.
Models are what we use to understand things, That's not how things are.