A Note on Principal Point Estimability