The problem with Solyndra wasn't only trying to scale up CIGS to mass fabrication (apparently hard to do) but also their insistence of doing it inside glass tubes (apparently much harder). All other CIGS companies have planar designs and as do this one. That's more promising as I see it.
Although I don't see the point of this project. Solar cells have to be cheap and I don't see how a CMOS compatible many step, large area, process will ever be. Seems like a better idea to just take a standard chip and power it with a standard solar cell than doing this, from a cost perspective.
Although, it's a research paper so I guess the point is to show more what's techically possible than what makes sense. (With the disclaimer that I haven't read the paper myself more than the summary above!)