I replied to one of Paul's Octave Records DSD misinfomercials describing the probable path of a DSD master through mixing, editing, layering, processing, prep for mastering, mastering, and distribution. Either he blocked it or nobody had a reply. I assume his fans hadn't a clue what I was talking about. Here's the clue: you can't edit, mix, or do any of the other things studios and producers (and artists demand) without taking the files back through PCM and/or even analog formats.
DSD was intended and designed for archival storage of audio material, where editing consists of start and stop. Despite nearly unlimited resources, Sony abandoned the DSD methods because the native files are unworkable and unwieldy without changing, recoding, transcoding, and converting to other workable formats. (Try telling a conductor that he must take it from the top because the 2nd oboe hit a blue note in the third movement of a Beethoven symphony.)
There might have been some arguments in favor of DSD quality back a few years ago when the SOTA in PCM was 16/48k. Today, we are doing 32-bit floating-point at rates of 256k+ samples, and quality in the recording is no longer a "nut to crack."
Today, production is a problem for music listeners -- the willful distortion of even good music to sound appealing on a ghetto blaster, Bluetooth speaker, Apple AirPods, or in a Chevy Malibu. Try cracking that nut...