Okay, looking at the paper more carefully (I'd just looked at the article before, and was arguing from my knowledge of the field more generally), I take it back -- there is definitely some formal category theory here.