If the authentication database is being breached, presumably the 2FA shared secret is going to be in the same database. Constructing the 2FA code would be trivial. After all, the server needs a way to check that a given 2FA code is correct for an authenticating user, so there has to be some way to generate those as well.

It would help protect against any other sites that you use the same password on, with a different 2FA shared secret.