The question is how hard is it to get experimental evidence for the thing.

If it occurs under a wide variety of circumstances at a high rate, it's easy.

If it involves an extremely unlikely set of circumstances, it's not easy, and will be hard to replicate.