The customer for this L402 protocol is obviously a media network, especially ebooks, and plausibly an alternative to spotify, applemusic and closer to the bandcamp model, content providers of all kinds who are currently subject to the chokepoint of payment processors. Emerging use cases would be domain specific LLMs trained on proprietary data sets in health, legal, accounting, and other professions. What it needs is a new there there on the other side of it.
The objection to PoW is thinly veiled objection to tying it to physical value instead of policy. The objection would be the same for gold. What they object to is physical constraints on political authorities, and these sort of "bitcoin bad!" comments should not be admitted to spoil serious discourse.
The economic case for PoW is that it converts any source of energy anywhere in the solar system to transferrable stored value. It would make a lot of space based energy harvesting projects economically viable, but also remote hydro projects that can't transmit electricity long distances, but can absolutely provide compute as an economic service from long distances and make the a local hydro system economically viable. PoW tokenizes electricity via compute. It's a very elegant idea, and the opposite of environmentally unsustainable because it actually rationalizes the price of electricity with endogenous demand.
The same people objecting to bitcoin micropayments schemes also object to nuclear because it would actually solve the zero-sum problem narratives they want to sustain and manage. These views are not without controversy, and personally I reject them on principle.