We obviously need it as an option to make the protocol censorship resistant.
Discussion
Doesn't affect the protocols censorship resistance at all where keys are stored, just makes the keys more secure.
Nostr as a protocol, uses the priv key to sign events very very frequently tho and the introduction of an external device for that is very inconvenient, nostr nsec's also do not have monetary value like btc private keys.
The signing devices will also have additional functionality introduced, so the bitcoin private keys in them are now more at risk of security vulnerabilities.
Overall not worth the trade-off imo, but if you want to do it more power to you, I just won't be using any hardware wallets that add nostr functionality, I want it doing only 1 thing - securing my btc keys.
The creation & storage of Nostr PKs are paramount to the concept of censorship resistance.
Just like your money, access to your online identity/communications can be critical;
Luckily, we already have backed up encryption models in plain English that can be dual purpose if necessary - segregated also if necessary - even with a simple change of passphrase (13th/25th word).
This concept is for LT storage of the Nostr PK - not signing every event - that's unnecessary.
Hopefully we can also then branch a master key that takes a dynamic approach with Nostr PKs - changing a derived nsec every so often - think of a longer term 2fauthenticator.
I would love to move about with my freedom toolkit stored in my brain with only 13-14words.