The cost of subsidizing solar PV installed in 2016 depends on the FIT rate that year. Right now the rate varies by system size between 0.135-0.221 euros/kWh (US$0.17-0.28/kWh). Following ammendments made to the EEG law in 2012, the degression rate for the FIT fluctuates by total new installed capacity — it can decrease by as much as 2.8%/mo if certain deployment targets or reached, or it can increase by 0.5% if fewer than 1000MW are installed on an annualized basis. Assuming FIT rates drop by about half by 2016, a more than reasonable assumption IMO, the cost of subsidizing solar installations will vary between about US$0.09-0.14/kWh — still more costly than the Olkiluoto reactor, but not by nearly as large a margin as the cumulative cost of subsidizing panels installed 2000-2011. And let's not forget that the EPR has an expected lifetime of 60 years, compared to a warranteed lifetime for PV panels of 25-30 years.
Of course as solar penetration rises the cumulative cost of the EEG law rises along with it. This year the German government increases the EEG renewables surcharge — the electricity tariff paid by households to cover the cost of subsidizing renewables — to 5.3 €cents/kWh, or US$0.068/kWh. The EEG surcharge now makes up the largest non-market tariff on German electricity bills. German electricity prices have already doubled since 2000.
Another ammendment to the EEG law was a cap on the total capacity of solar that can be subsidized by the FIT at 52 GW. This shouldn't be that surprising given the massive swings in output from solar PV in Germany. On that famed sunny Saturday afternoon last May, the roughly 30 GW of German solar generated about half of the nation's electricity — but total annual solar generation was less than 5% German electricity. So if total deployment were to double, solar would generate only about 10% of total electricity but often run the risk of generating >100% of load and overwhelming the grid. New provisions are being put in place to allow compensated curtailment whereby PV owners/operators will receive 95% of the FIT rate for curtailing their solar generation when the grid doesn't need it, at which point the German government will be paying people not to generate electricity on a regular basis. The exact costs and consequences of this evolving system are yet to be seen, but they will certainly be instructive and likely evident by the end of the decade.
Finally, a few notes about motivations.
The German feed-in tariff is a national perogative predicated not just on decarbonization, but on renewables innovation, industrial policy, and the decentralization of power supply. These are all worthy goals that the German populous has thus far proven perfectly willing to pay for, and good for them. It is indeed likely the case that the world would not have seen dramatic cost reductions in solar technology in the last few years without Germany's FIT, the US's ITC, and other national solar subsidy policies. While China's powerful industrial policy, subsidized production, and likely dumping of solar products onto world markets played the biggest role in declining module costs, it is unlikely China would have pursued these efforts lacking subsidized demand in mostly Western economies.
But these motivations are in contrast to a regular argument made that the German solar story proves that solar PV is cheap, scalable, and that solar negates the need for more costly technologies like nuclear power or other centralized forms of power production. This is simply not the case yet. In the world's most accomplished solar regime, Germany, solar provides less than 5% of electricity after a decade of deployment, electricity costs continue to rise, carbon emissions are expected to stay flat through 2020, and it's unclear how the grid will continue to deploy solar after a few more years as intermittency forces mandated curtailment policies and FIT incentives cease. The case for renewables is certainly improved by accounting for growing wind and biomass in Germany, but these two bring their own complications in the form of intermittency (in the case of wind) and public acceptance and land use implications (in the case of both).
Germany has done more than almost any other country to develop renewables, grow clean energy markets, and bring down technology and soft costs. But to insist that the German model can be exported to the rest of the world as a coherent and comprehensive strategy for decarbonization is a dangerous stretch of reality. Global energy demand will triple or quadruple by the end of the century if we're lucky, as billions of people lift themselves out of withering energy poverty. To suggest that their needs can be met by renewables alone, and that technologies like nuclear power are not needed, while renewables are still carving out niches in subsidized markets, is foolhardy at best and dangerous at worst.