The whole premise is bad. If the supervisor can do it in 2 months, then they can do it in 2 weeks with AI.

Didn't PhD projects used to be about advancing the state of art?

Maybe we'll get back to that.