This was a great reply, thanks, yo :). Overwhelmingly agree. Maybe more thoughts later when I'm not on mobile. The only thing I'd like to characterize better is "energy" costs to manufacture e.g. paper vs. costs to maintain NFT blockchain records. Could be bad for blockchains. Proof of stake NFTs? Dunno. I need to think more about that. Also, I learned early that life is shit by spending my allowance on Star Wars customizable card game cards. They're worthless now.
So... look at a real use case - Arianee. Arianee builds a framework for Breitling (among others) to verify their product via NFT. This is a private blockchain (Ethereum white-label) where the only write-points are controlled by Breitling and Arianee. They're readable from anywhere - but the robustness of the network only matters to Breitling and its customers so it doesn't take many. When I talked to Arianee they said they have about sixty nodes. that's everyone on Arianee, so Vacheron, Panerai, IBM, etc. Mining performance really doesn't matter because they aren't mining, so you can run that stuff on a Raspberry Pi, or as a tiny subprocess on a virtual server. That's proof-of-work, by the way, but proof-of-stake is the same because you're just there, you aren't trying to get paid. Sixty nodes at 15 watts is 900W, for every NFT on Arianee. Breitling puts out 200,000 watches a year, every year, and has been on Arianee since 2017. A sheet of paper, by way of comparison, is 50W/h. That's before you print it, of course, or package it, or distribute it. And that's just Breitling.The only thing I'd like to characterize better is "energy" costs to manufacture e.g. paper vs. costs to maintain NFT blockchain records.