>If you're promoting research into quantum gravity that doesn't respect special relativity, you're a crackpot.

I think this statement is off the mark. Theoretical physics is often about simplifying a problem to a form that you know how to solve, even if it's not physically realistic. Once you have a solution in your toy domain, you might have a chance of modifying it to find a physically viable theory. A solution can also be useful in restricted domains, even if it's not true in the general case.

If we look at the historical development of quantum theory, it wasn't modified to take relativity into account until 1928. Even then, the non-relativistic theory has independent value and makes many valid predictions.

Another example is the AdS/CFT correspondence. We know we don't live in anti-de Sitter space, but study of it has led to a better understanding of black holes.

I don't think we should rule out any line of research a priori. We should throw a bunch of different approaches at the problem and see which ones help us to make progress.

Of course, crackpots are a real thing. In my mind, the definition of a crackpot is someone who claims to have a revolutionary theory without any track record of engaging with research-level ideas. Perusing Sabine's Google Scholar page shows that she has a respectable number of papers and citations, so it is worth taking her ideas seriously, even if you disagree with them.

Can you make a cogent argument against one bit of her research instead of calling her a crackpot for challenging things?

Overall, she comes across as a very clear communicator with criticisms (rightly or wrongly) of some current research.

Is there a short argument for why MOND cannot agree with relativity, at least on the relevant scales?