Comment on "A new method for the nonlinear transformation of means and covariances in filters and estimators" [with authors' reply]