Most of this is noise; a proof does not 'create' the equivalence of sensitivities, it demonstrates it. What we need to control are the conditions and limits of this equivalence, such as the regularity of functions. For example, in complex financial environments, convexity assumptions may not hold, making the equivalence less relevant for managing pension fund investments.