Some Remarks on a Generalization of the Superintegrable Chiral Potts Model