Some failure states are unique to people who exist in these weird edge-case states though. Like the person who had their luggage stolen, the person registered the laptop to their own account, then returned it still paired. And apple wouldn’t un-pair it from Find My even with a police report documenting it all, therefore it’s bricked.
(And to be fair to apple here - they didn’t do anything wrong here, strong end-to-end security inherently means allowing these states. Otherwise the cops could order apple to unlock it too, and apple wouldn’t have a moral ground to object if they’re regularly performing the task in other circumstances. Otherwise people could social-engineer apple support to unlock a stolen device, or their partners. To a certain mindset, google and apple not having any real support is a strength because there’s no way to social-engineer your way past the actual security. But people want both the idea of E2E security and the convenience of being able to remotely un-register a laptop from someone else's account...)
Anyway, that failure mode wouldn’t exist if they were logged in to their account, and e2e encryption makes that a very low-risk thing overall.
Apple can’t see where to it devices are anyway, without doing a song-and-dance to authorize the session on a pre-authed device. Airtags and iphones have a rolling hardware identifier for bluetooth and wifi based on a cryptographically strong pseudorandom sequence, and apple can't correlate the identifiers back to an actual device without a pre-authed device relaying the sequence from your account. Etc etc.
Apple have actually done the legwork to make sure they can't see anything (or be forced to reveal anything) if you don't want them to (by enabling E2E), and that actually does drive a lot of "user-unfriendly decisions". And sure, android people will say "that's awfully convenient", but, the end state is still a lot stronger than any other major offering regardless of why you think they're doing it.
(And to be fair to apple here - they didn’t do anything wrong here, strong end-to-end security inherently means allowing these states. Otherwise the cops could order apple to unlock it too, and apple wouldn’t have a moral ground to object if they’re regularly performing the task in other circumstances. Otherwise people could social-engineer apple support to unlock a stolen device, or their partners. To a certain mindset, google and apple not having any real support is a strength because there’s no way to social-engineer your way past the actual security. But people want both the idea of E2E security and the convenience of being able to remotely un-register a laptop from someone else's account...)
Anyway, that failure mode wouldn’t exist if they were logged in to their account, and e2e encryption makes that a very low-risk thing overall.
Apple can’t see where to it devices are anyway, without doing a song-and-dance to authorize the session on a pre-authed device. Airtags and iphones have a rolling hardware identifier for bluetooth and wifi based on a cryptographically strong pseudorandom sequence, and apple can't correlate the identifiers back to an actual device without a pre-authed device relaying the sequence from your account. Etc etc.
Apple have actually done the legwork to make sure they can't see anything (or be forced to reveal anything) if you don't want them to (by enabling E2E), and that actually does drive a lot of "user-unfriendly decisions". And sure, android people will say "that's awfully convenient", but, the end state is still a lot stronger than any other major offering regardless of why you think they're doing it.