Model distraction: A comment on Daveri and Tabellini