In my field we don't publish with PhD students, that is bad for their carreers :-)

You create a public link (ie everyone can access it) for a given folder, then define roles (can add/edit/delete files).

CNRS Sdrive (based on Nextcloud) is just perfectly making its job on sovereign drives and OSS tools. And now (in France) we have Visio that is perfect to replace GoogleMeet or Zoom.

Did you also consider that, at least at MDPI, guest editors papers might be APC-waived? polecopub.hypotheses.org/3231

Yes, I did (a chapter in an edited book), in order to see how things go. I just received a confirmation email for the moment.

They are not a "leading scientific journal" but a newsmagazine which also happens to publish academic articles.

Very classical capture by economic/political interests. Criticizing/debunking means more work instead of copy/paste/transcript and less short-term opportunities (interviews, invitations..). Every Trump/Macron/... declaration should also be met with skepticism, they usually don't.

In a world where Els. and others close to extract more value (see abstract and metadata in Crossref), it is totally relevant. The burning question is probably how to block, manage, and sort the deluge of fake data (papers, refs) now easily produced by papermills and kindly published by legacy pub.

That has happened many times the other way around: reviewer asking for co-authorship or should I say, demanding, to show his (more than her) appreciation of the manuscript.

My point: above and beyond the burning of money and coal, these things just don't work (except for cheaters and criminals), there is no need for nuance.