Here's the kicker: The Einsteinian model of curved spacetime is extremely good at predicting what we actually see when we observe objects in the neighbourhood of large masses. and that's what matters.
Because here is the thing that river is missing: when we make theories - such as Einstein's general relativity - we are not trying to describe what reality "is". That is not what science is about. Science is about making predictive models. His model predicts what we see exceedingly well - far better than any other model produced.
He is not wrong trying to find better alternative.
I am on board. Our understanding of gravity can improve, hopefully to the point where we can manipulate it.
Yet, pointing out incompleteness of our current understanding do not validate alternative that is proposed in consequence.
If he could let go of attacking what is, and actually try to present what he is proposing he would have a fair shot on testing his idea.
As for now, I do not see how his idea could offer anything even close to what is.
So, he is not wrong thinking there have to be a better way to model and understand gravity, he is wrong (my opinion) that what he is proposing is actually it.
Nature of novel ideas is that 99,999% of them are doomed to fail.
So, being wrong takes nothing away from River.
Not testing your idea, not being able to accept it has failed and move on to next one. Is the problem.
Because you can get stuck with it.
For example:
It takes Venus longer to rotate once on its axis than to complete one orbit of the Sun. That's 243 Earth days.
Venus has a surface gravity that is about 91% of Earth's gravity.
Venus has a mass that is about 81.5% of Earth's mass.
So, similar mass, similar gravity, yet totally different rotation speed.
If gravity would be rotation related, should it not reflect on Venus gravity?