Nah he's out of touch with modern research. Publishers are flooded with high quality human-made manuscripts, not just AI crap. This is especially true of math and physics. Submission have increased far in excess of audience size or # of publishers, so higher rejection rates are inevitable. In math or physics at least, desk rejects are more common due to the editors being more picky due to huge volume, not that the quality of research is worse.