I agree with your points but they don't really support "luck isn't random". I would say the argument is more "if you increase luck surface area the expected outcome will be higher". Not as catchy.