The market of suppliers will grow to meet the consumer demand, and suppliers will invest some of their profits into R&D and manufacturing. And I believe Moore's Law applies to the manufacturing of solar cells.
http://spectrum.ieee.org/tech-talk/semiconductors/devices/ph...
Create a market and let private industry handle the rest. Space, robotics, cheap energy, etc.
For instance, these 35+ % efficient cells have been available for two years now:
http://gizmodo.com/5388795/sharp-triple-layer-solar-cell-set...
But you wouldn't be putting them on your roof unless you wanted to achieve a specific number of Watts on a small surface.
The figure to monitor is $/Watt, and that's where any gains will come from. Any decrease in the price per Watt is real, immediate gain. Increases in efficiency are only good if they lead to a lower price per Watt.
Surely they know enough now about thin-cell solar to invest into more R&D for it. Give the 75 million to 75 professors to have 5 new PhD students each and a bag of money for experiments. Given the history so far[1] I don't think a few percentage points more from 375 PhDs is out of the question. Those few points would have a far more dramatic effect in the long-run than just helping to finance installation of current technology for the equivalent of one small town.
[1]http://upload.wikimedia.org/wikipedia/commons/e/ed/PVeff%28r...