then the evidence required may far exceed what any one person could gather

Yes, it can.

But it's possible to do much better on in these situations than people usually do, by developing good heuristics for how to weight the beliefs of others who have relevant subject matter knowledge. If you haven't seen them, see e.g. Vladimir M's post Some Heuristics for Evaluating the Soundness of the Academic Mainstream in Unfamiliar Fields and Nick Beckstead's post Common sense as a prior.