Sure, posting under a well-known research institutions own domain is a strong link in the trust chain in this example.

Of course you have to be able to trust that some post you find (which names a real person doing work in the field) really comes from that person and is not simply stealing their authority.

My point wasn't that this specific case is like that, but rather that online spaces are bombarded with claims that take far less effort to generate than it takes to verify or debunk. And once some story is seen by enough eyeballs, no withdrawal or retraction is going to have the same reach (bad actors already know this).