Interestingly, my father, a moderately respected scientist, has cited similar reasoning to me when discussing why he believes in supernatural phenomena. He believes he has encountered overwhelmingly convincing evidence, but says he understands that I shouldn't necessarily believe him. This is... a pleasant way to deal with disagreement, if not faultless reasoning.

After reading your thread with gwern, I think you and he are probably wrong about this reasoning in general, and you are probably wrong in your case specifically.

I think it should be possible to encounter supernatural phenomena in such a way that it is extremely convincing to you and not to anyone else. If you were a highly rational agent who encountered real supernatural phenomena, and told (even perfectly rational) people about it, their first reaction would be not to believe you. And this likely makes sense on their part unless you're able to produce extremely good evidence that (you are highly rational AND you are very unlikely to be lying), OR you have reproducible evidence of a particular phenomenon that you can show them.

But you should be able to produce such evidence... if it's not convincing to them, why is it convincing to you?