The Scope, Limits and Distinctiveness of the Method of "Deduction from the Phenomena": Some Lessons from Newton's "Demonstrations" in Optics