Excellent series. I'm in strong agreement with the bulk of it. I continue to think, though, that you are overstating the role of R&D relative to deployment. I reviewed the key Nemet reference you cite for the proposition that cost declines are the result of "equal parts ongoing R&D and economies of scale" and do not believe it supports that conclusion.
Rather, Nemet concludes that 30 percent of the total price decline of PV is from improving cell efficiencies, of which R&D is the primary driver, although learning by doing is credited with "some" contribution to increasing cell efficiency as well. So R&D contributes roughly one-quarter of the overall price decline, rather than half.
More than half the decline is directly attributed to factors related to growth in the market for PV cells, especially larger manufacturing plant sizes (43%), but also learning by doing, standardization, market predictability, competition, etc. These are all factors that deployment policies, such as renewable standards and incentives, aim to effect.
Nemet attributes 12-15% of the price decline to spillover effects from the IT industry in reducing silicon price and usage. But it is interesting to note that around the time the paper was written (2005) was when PV price decreases halted as a result of a silicon shortage. Up to that time, the PV industry largely utilized silicon waste from the IT industry. The huge decline in PV price since the paper was written has been largely driven by the creation of large dedicated silicon plants to serve the PV industry. Those plants, of course, were enabled by the large expansion of the PV market--driven by deployment policies.
Even within the R&D contribution, Nemet finds that industry was responsible for 6 of 16 key breakthroughs. The other 10 breakthroughs were from universities and government research. The increasing scale of the industry likely contributed to industry's ability to conduct such R&D, and I would expect industry to play an ever larger R&D role as its scale increases.
While Nemet looks only at PV, he also cites a paper (Madson et. al.) that found that scale made an even larger direct contribution (60%) to price reductions for wind energy, vs. 43% in direct scale effects Nemet found for PV.
Finally, the specific example you provide to illustrate the high cost of deploying with yesterday's technology makes no sense. It is not meaningful in any way to multiply 30 year old costs times a target penetration rate to calculate the cost of deployment. Every technology has to be ramped up over time, and in that process of ramping up costs will decline due to scale, learning, etc. That is the whole point.
Bottom line: yes, government and university R&D has a very important role to play in the innovation process. I strongly agree with the many recommendations that it should be at least doubled. But it is significantly overstating the case to assign it an equal or dominant role in the policy portfolio.
Clean Energy Consulting