This would be even better if it was possible to annotate the randomly generated "proof" description with links to any actual proofs following that skeletal description.
Makes me think of an article I read many years ago (in, I think, Mathematics Magazine) titled "Confusion Rings" which parodied abstract algebra papers. About all I remember now is "heterocococomorphism".
My roommate wrote this last year as a joke for CMU students. It's a simple script that pulls from a pre selected discrete set.
Author is Alan Pierce btw.