Comments on "Lower Bounds to Error Probability in the Presence of Large Intersymbol Interference"