re: The point was about the economics at the next two nodes.
In economics the question is whether people will buy it. And the answer is quite clearly yes as TSMC already has a whole pile of customers lined up for 16nm in 2015.
I really doubt there is anything Intel can do that others can't match, the main question is how quickly they can do it, i.e. when not if. If TSMC really do deliver on 10nm at double the density compared to 16nm with risk production in 2016 and volume in 2017 like they seem to be saying are we really certain it won't be economic?