What type of threat scenario does this address?
If a password dump of ServiceA leaks, and it contains the credentials for a user, and this user have an account at ServiceB protected by 2FA, nobody will be able to authenticate as him, even with the correct password. As password reuse is a common problem with credentials (the vast majority of users have few passwords and use the same passwords on every service).
After all, an attacker who would be able to authenticate as a legitimate user would then be able to change the password and lock the legitimate user out, so what is the difference?
If the account is protected by MFA, an attacker will only be able to authenticate as the user if the MFA token was stolen too. If the attacker have the password but not the MFA, those procedures would help protect the account in the case of a password leak.
If, for example, the legitimate user lose his phone with the MFA token, or his hardware token, but knows the password, he can prove ownership of the account by supplying his ID or another details proving he is the owner, before disabling the 2FA and allowing him to login again.