Isn't the largest labor input for solar in manufacturing and installation, rather than maintenance? Using your example of 12,346 GWh of solar PV for 2012, wouldn't those solar installations have produced the same amount of power (or close to it) in 2013 without any additional labor input, plus whatever new capacity was installed in 2013, resulting a higher GWh:worker ratio?
Home solar installation costs are too high. Rocky Mountain Institute, 2011: "Total soft costs—including customer acquisition; installation labor; permitting, inspection, and interconnection (PII); and margin and other associated costs—now make up approximately 70% of the total installed priced for a U.S. residential PV system."
Australia and Germany have lower costs per KW installed, but the installation cost exceeds panel cost even there. This is a big problem. Even if panel cost goes to 0, installed cost will be maybe half what it is now.
Solar panels which are also roof components have been suggested for new construction, but that requires more roof design standardization than the housing industry is used to.
There are also "off-roof electrical costs". A solar installation usually requires a new meter box and additional boxes for inverters, control gear, and disconnects. Standardizing how solar systems connect at the meter box would simplify that.
The norm in Australia is now a one-day install. The US and Germany run slower, partly because more waterproofing is required. ("Every hole in the roof is a dagger through the heart of a house")
Good data, and yes, RMI have done some studies on this.
"Even if panel cost goes to 0, installed cost will be maybe half what it is now."
I've drawn the relationship between this and Amdahl's law, in parallel computing. As parallelization increases, it's the serial component of processing which comes to dominate. Eventually you reach a point where further parallelization yields no benefit.
I'm curious about your username. I know of animats from Carl Zimmer's article on them a ways back, and found the concepts of complexity and capabilities in evolution fascinating.
