The point of my definition, though, is to make determining whether an object is a planet actually useful for our understanding of the formation of the solar system, rather than merely a stamp-collecting exercise.

As to your example, if I found a rocky, Mercury-sized object in the Oort cloud (which I think is what you meant, as Kuiper belt objects are thought to have formed in the proplyd), I'd have to conclude that it wasn't a comet, just the same as if I found a brown dwarf in the Oort cloud I'd have to conclude that it wasn't a comet.