I think there is utility in having non-metabolic-ward studies, but for the primary purpose of highlighting sociological issues surrounding diet.

You need the ward, of course, to prove what works and what doesn't under perfect conditions. But you also have to figure out how to export those conditions from the ward.