I'm a bit surprised by a wafer pricing chart. Sure, the individual wafer cost goes up; but if you have a design with the same transistor count, the size of your chip is going _down_, which means that you'll have _more_ chips per wafer (assuming comparable yields), which means that individual chip cost will go down.
(40/28)^2 ~ 2, so if 28nm is not that bad comparing to 40nm, assuming, again, comparable yields.
This is precisely the analysis given in the per-transistor pricing crossover chart. That shows that the old rate of wafer price increase still allowed for significant per-transistor cost advantage fairly quickly. The whole point is that the rate of wafer price increase has itself increased to the point where these crossovers are barely happening anymore.
(40/28)^2 ~ 2, so if 28nm is not that bad comparing to 40nm, assuming, again, comparable yields.