RomeoStevens comments on The robust beauty of improper linear models