At first glance, good news and bad news.
Good news: this whole DMA terminology bothered me with the (pardon the word but it is valid in this context) co-mingling of DMA and CPU explanations. TPL is trying to clarify and rectify any ambiguity.
Bad news: TPL is essentially acknowledging that they didn't do the fullest due diligence in their earlier research and presentations and now, a LONG time later, are scrambling to offset their oversight.
It'll be interesting to see how this turns out.