Gravar-mail: Commentary: The formal approach to quantitative causal inference in epidemiology: misguided or misrepresented?