100%

I've discussed this with fiatjaf, and he argues that nip-26 delegation token revocation can't be guaranteed;

I still think it should happen:

* nIp-26 delegate key to a new pubkey

* if pubkey abuses, delegator creates an opentimestamped event revoking and blasts the revocation to all known relays

* if delegatee continues to abuse delegator has timestamped proof that abuse was not authorized by the delegator