I guess the definition I would use to describe a "kind" would be a group of living organisms having descended from the same ancestral gene pool.
So what criteria do Creationists use to determine if two species are derived from the same ancestral pool? Do they use morphological criteria? Genetic?
Even more important: by what objective tests would a creationist determine that any two given species must have absolutely no ancestor in common? This particular claim (i.e., the existence of historically-disconnected pools of ancestry) is the very core of their problem; shedding light on it would go a long way towards building an actual theory of kinds.