"but maybe this could have happened"* is the exact evidence being input into your bayesian model. The fact that this particular thing could have happened is a bit surprising, so it does count as some evidence, but it's not strong enough to then go run around saying on the internet it definitely did happen. Especially in the context where, as you point out, there's also other evidence of the form "but maybe <this other thing> could have happened and that's also surprising".

Which to address a different subthread, is exactly what some people did (go around on the internet confidently stating it did happen). Which is why, I think, other people then labelled the people seriously discussing the theory as conspiracy theorists. Which is a step too far, it's only the people confidently asserting from weak evidence that it definitely did happen who should be labelled as conspiracy theorists. Which is all to say since when does the internet do any of this nuance at all well on any side.

* GP's phrasing, which is arguably different from your phrasing of "but maybe it's true"

JFC, finally someone that understands specific evidence changes probabilities.

It's unlikely we'll ever know the truth. If it's a cover up, then it's possible that someone will come forward in 30 years. For example, Luis Salas spilled the beans on Lyndon B Johnson's 1948 election fraud. But it's unlikely, because the PRC doesn't have a statute of limitations on STFU. If it's natural zoonosis, then maybe we'll manage to find and prove the origin.