> there's only guaranteed to be a maximal hom object if Obj×Obj is finite.

Good point, I'll update the statement of the theorem.