A few corrections / comments:
Panel lifetime is generally given as a nominal 20-25 years. That's not a minimum. The reason is a gradual degradation, mean of 0.8%, median 0.5% of output per year, due to multiple causes. See: "Photovoltaic Degradation Rates: An Analytical Review", Dirk C. Jordan and Sarah R. Kurtz, NREL
http://www.nrel.gov/docs/fy12osti/51664.pdf
I would like to know how the employment numbers here are calculated, and how they're allocated across the lifetime of solar plant. I'm not trying to knock solar, I'm showing that it's quite possible to have too much of a good thing though.
CleanTechnica is generally a pretty reliable source. US EIA gives 3.2 GW nameplate capacity for solar as of 2012: http://www.eia.gov/electricity/annual/html/epa_04_03.html Installations have been going at a major clip, so 4.8 GW sounds highly plausible.
"I think the calculation is more complicated than you make it seem."
It almost certainly is. However my point was that it's also a bit less straightforward than the BusinessInsider piece presents. Believe me, I'd like to see good jobs in the solar sector. But I'd really like for those to be productive jobs.
I do think that solar will continue to see cost reductions (and those will be key to its success). Addressing factors other than just efficiency (mostly overrated), and in particular, the installation, durability, and maintenance costs, are what I suspect are key.
Jobs ... are complicated. My point isn't so much that they're good or bad, more that if the trends reported are projected forward, we'll see a very large portion of the labor force directly employed in energy harvesting. This isn't the end of the world, but it does mean that roughly 1 in 5 present jobs will be supersceded. There was an earlier time when energy harvesting occupied a similar or larger portion of the workforce, though it was called "farming" at the time. In some countries it still comprises 90% or more of all work.
This plays into another discussion -- that of EROEI and what levels of it are required to support various levels of economic and technological complexity.