Do you also think that arXiv is too strict? Enter http://vixra.org for all your even crazier “papers” without any exclusive measures such as “editorial control” or “vetting” or “even the lowest bar for quality control” or similar ableist selection criteria.
Just among the latest 10 CS papers, I found a linear-time sorting algorithm, linear-time integer factorization, and proofs for both P != NP and P = NP (the latter used to derive an efficient algorithm for the halting problem), so you know it is full of good, revolutionary and very sane stuff.
It’s got all the other disciplines as well, so if you’re looking for physics or psychology or anything else, don’t hold back. 33 mentions of Atlantis on the Archeology overview, and two papers mentioning unobtanium (one might be a spelling mistake, the other seems to do so unironically). http://vixra.org/