Taleb mentioned on Twitter (with a draft paper to back it up) that the observed (ie. by Picketty) inequality increase could just be a statistical illusion given the methods of measuring it.