Critiquing bibliometrics with Goodhart’s Law