For a business, implementing FIDO in this case seems like a win. They get more lockout of password sharing, and have to explicitly work to enable shared logins (if that's something they want). Of course it's nice for an end user to be able to share credentials, but I think long term outlook will be towards businesses and applications pushing to improve their user management tooling.
Anecdotally, in the instance you described they net the same $60/yr regardless of wether you use your wife's login, so they're getting more stringent access controls (though you could always use the device in question, too) in line with their terms of service.
(Would also note that my response does not cover your latter points on helping users manage their accounts with their presence, but without their device presence, which I agree is a valid concern)
You also ignored the point that perhaps they'd have gotten to like the service enough to get a subscription for themselves, but refusing out of spite now.
Anecdotally, in the instance you described they net the same $60/yr regardless of wether you use your wife's login, so they're getting more stringent access controls (though you could always use the device in question, too) in line with their terms of service.
(Would also note that my response does not cover your latter points on helping users manage their accounts with their presence, but without their device presence, which I agree is a valid concern)