From Plane Crashes to Algorithmic Harm: Applicability of Safety Engineering Frameworks for Responsible ML