Ericf comments on The reverse Goodhart problem