> "Undefined behavior" is not a meaningless made up term that you can redefine at will.

Sure, I agree with that.

> The word "undefined" has a clear meaning: there is no behavior defined at all for what a given piece of code will do...

That is true, but...

> ...meaning it can literally do anything.

This is not at all true! That is a different (but closely related) matter, which is "what is to be done about undefined behavior". Which is certainly something one has to take a stance on when working to a language spec that has undefined behavior, but that does not mean that "undefined" automatically means your preferred interpretation of how to handle undefined behavior.

The original question is how UB is defined, not about the preferred way of dealing with it in a practical sense. And the definition of UB is behavior for which the language definition imposes no requirements, and explicitly leaves open the possibility of ignoring the situation altogether with unpredictable results.