Author reputation and citation patterns provide plenty(?) of signal without journal/reviewer/editor endorsement. But you could still imagine introducing “badges” that provide similar additional signal to what publication in a top journal provides today. Academic societies issue a fixed number of badges to top preprints each year. But ditch the song and dance around peer review.