Vasco Grilo comments on When should the inverse-variance method be applied to distributions?