[explanibrag]

Yes, but my original point was that the foundational model isn't convincing at all to the layman. It can't stand on its own feet, and needs additional structure. Any decent model should come with caveats, and should specify the domain in which it's applicable.

Krugman was using the model to argue in favour of free trade. He'd be much better off arguing using historical examples if he wanted to be convincing.