Good article, but you have omitted or simplified some of the technical factors that might be viewed as externalities, but are still important. There's a good article in this month's IEEE Power and Energy magazine (unfortunately behind a firewall at http://www.ieee.org/portal/site/pes/) which goes into some detail on these issues, particularly as they relate to the interaction between various energy sources. A few of the relevant points they make include:

• Time-of-day relationship between wind availability, solar availability, rotating generators, and demand. Solar supply usually peaks at noon, wind is skewed towards the afternoon, and demand peaks in the late afternoon, so you may still need to provide conventional power with its associated fixed costs. Or batteries, with several cost implications.

• Similarly, if you add distributed renewable resources onto the distribution grid, the direction of power flow may change depending on the time of day. This causes all sorts of new problems in the area of voltage control, frequency control, and reactive power control. The fixes for these problems are not even very well understood from the theoretical viewpoint, and the required technology is still in development.

• At the local level, if you plan to use distributed storage (for example, batteries in plug-in hybrid or electric cars) as supply leveling sources, then the neighborhood distribution grid has to be designed with considerable care. For example, at some times of day the power may mostly be flowing from the cars into the grid (to supply power for lighting), which means that the voltage gradient through the neighborhood is the opposite of what it normally would be.

Bottom line is that as the fraction of power supplied by non-traditional sources increases, the need for changes to the overall system design also increase.

None of this means that alternative energy is bad, on the contrary, it is obviously good that practical sustainable supplies are economically viable. What is needed, in addition to further investment in the technology at the source (wind turbines, PV, etc.), but also considerable attention to the regulatory environment that guides grid development and also to the technology that will support the future grid infrastructure.

Here's a link to a neighborhood-scale PV project in Freiburg, Germany that you are undoubtedly aware of; the power distribution system in that neighborhood was of a new design. As an example of the practical difficulties encountered, (quoting from the IEEE article): "Another interesting effect occurred in some of the residences where more frequent overvoltage disconnects were reported. Eventually, it was found that the connection from the inverters to the utility connection switchboard had been made with regular 1.5 mm2 wire [sic, probably 15 mm2] [instead of the specified 35 mm2 wire]. This wire size is sufficient for the expected current from the inverter, but too small to keep the voltage drop low enough."
http://www.werkstatt-stadt.de/en/projects/22/

While particular this example is specific to distributed PV, it is a related subset of the general problem of sustainable electricity supply. The cost of changing from wind to PV or whatever includes not only the cost at the source (including financing cost), but also the distribution cost and the costs of integrating it into a comprehensive system. As the case above shows, for example, you have a whole community of electricity installation tradesmen that has to learn new ways of doing things--obviously a costly proposal.

Which I think is what you are saying in your main article...