And that is very solid bayesian logic. Wuhan has the only BSL-4 lab in China, and was specifically working on gain-of-function research on coronaviruses.
The initial prior of zoonotic origin simply because that was usually the case in the past is just as circumstantial, but also just as solid in bayesian terms.
>>nwah1+(OP)
Belief in the posterior update is still completely circumstantial. It's not direct evidence, nor an explained cause other than proximity. I'd argue they miscalcualted the probability change with respect to this piece of information.