In theory it should work, but I'm still suspicious. In general, thicc compatibility layers seem to work in practice, and everything else never really seems to reach IBM PC levels of compatibility, fragmentation seems to happen fast if there's no consensus driving everything to follow some One True Standard.
We'd probably wind up with 5 UEFI alikes, 3 distros that don't use any of them, and there would probably be subtle compatibility issues.
Eventually the consensus would evolve to some new standard, the way systemd has, and then everyone would complain about that being too heavy and how they don't want a monoculture.
Or, worse, we could wind up in the same place ARM is, for a decade.
There's no technical reason for it, but there also doesn't seem to be a very strong pressure to standardize.
Consumers aren't exactly going to unionize and engineers love to fragment, so to me a standard with that level of popularity is something special and should only be messed with if you have a real clear plan to preserve compatibility.
We'd probably wind up with 5 UEFI alikes, 3 distros that don't use any of them, and there would probably be subtle compatibility issues.
Eventually the consensus would evolve to some new standard, the way systemd has, and then everyone would complain about that being too heavy and how they don't want a monoculture.
Or, worse, we could wind up in the same place ARM is, for a decade.
There's no technical reason for it, but there also doesn't seem to be a very strong pressure to standardize.
Consumers aren't exactly going to unionize and engineers love to fragment, so to me a standard with that level of popularity is something special and should only be messed with if you have a real clear plan to preserve compatibility.