I guess that doesn't apply to the NYJM, but I didn't ask for them to be the journal doing the publishing (just noted that their approach of dealing with the fallout of Rivin's machinations was unfortunate).
If push came to shove and I'd be on the board of editors of a publication of same or higher posture than an affected publication with overlapping audience, I hope I'd at least consider doing that (but as mentioned: not through subterfugue, but officially).
That's political, but in a way it's merely self-preservation: Niemöller's poem "First they came ..." and all that.
But I disagree with you that a real journal has a duty to maintain a bad-faith off-topic out-of-process submission. That's a disservice to the cite record.
So completely removing a paper that was published in violation of process is a process violation itself, undermining the reliability of citations. At least they replaced it by a paper of the same length, hopefully preventing any following papers from moving around, but they really shouldn't reuse space like that. Especially considering it's an online journal where pages are essentially free.