Remarks on a paper by Skornyakov concerning rings for which every module is a direct sum of left ideals