Man, I have half of a paper written on exactly this topic, arguing for a similar conclusion. I frame it by saying that there are lots of different concepts or meanings that can “do business” under the heading of a single word, and often, ordinary usage doesn’t distinguish between them, because they coincide. But when we aim to offer a definition, we end up picking out one of the meanings, but the others are there to generate counterexamples.

Ignoring this generates lots of philosophical confusion. This is really obvious in debates over the definition of consciousness, but also, somewhat ironically, seems to happen constantly in philosophy of language. Think, for instance, of debates over the nature of lying or dog whistling or implicating or so many others. Or the debate over manipulation in ethics. It seems clear to me that there are just several different concepts in play each time, and analysis in terms of N&S conditions is hopeless.

All that is to say: great post. Totally agree.