I thought they were just bolstering the refutation of TFA's assertion that XHTML is strictly better because of its parsing algorithm.