The way you usually know is via peer review -- e.g. other people previously declared to have produced good things declare that MIRI produces good things.

I think this isn't really cutting to the heart of things--which seems to be 'reputation among intellectuals,' which is related to 'reputation among academia,' which is related to 'journal articles survive the peer review process.' It seems to me that the peer review process as it exists now is a pretty terrible way of capturing reputation among intellectuals, and that we could do something considerably better with the technology we have now.