Comments on "FF-Pade method of model reduction in frequency domain" [with reply]