Scope Change -- Worth It?

I recently heard an interview with Kent Alstad in which he commented something to this effect: "If a change in scope is proposed for a project, ask yourself this: will the project succeed if we do NOT do this? If the answer is yes, don't do it."

That's about as clean a way to judge whether to give in when scope begins to creep (as it always does) as I've ever heard. The fixed price guys need to mind this. The agile guys will have a nice counter to the whole premise of scope, I'm sure.