Capital-energy substitutability in manufacturing sectors: methodological and policy implications