> Alike, most online journals remove stuff by wholly replacing it with a retraction notice
Can you provide a copy, or even examples? I can pull up dozens of Science, Nature, and Physical Review retracted articles to support my point. If those huge families of journals behave as I say, on what basis are you claiming "most" journal don't?
> What no journal does is to replace erroneously published article with another entirely unrelated one in-place.
Did anyone ever suggest this was the case?
> the ArXiv does have a moderation team and stuff is being removed
I know several members of the moderation team. They of course reject some articles at the time of submission, but I'm not aware of any being retracted and un-published after being posted publically. Can you please link to such a retraction, or to evidence of this in the Internet Archive?