Pamela, great point - and I am not arguing that $/watt should be the only (or even primary) metric. LCOE is always most important.
My goal is simply to isolate a single variable, to understand how it plays off against different applications. But there are many, many other forces. Besides de-lamination, these technologies have very different thermal coefficients, low-light absorption, widely varying voltages (which drive string sizing and copper length), and different racking & grounding requirements. I don't mean to downplay these. Just 'abstracting them out' to isolate on $/watt vs. conversion efficiency.
Hi, sorry for the delay - was on the road.
AWB: I don't have good info on fossil fuel economics, but I'll go back to one of my favorite textbooks: Renewable and Efficient Electric Power Systems, by Gil Masters. He has some budgetary values for the CAPEX and fuel costs for coal/gas/nuclear. I'll look them up when I'm back in the office (mid-week) and post them.
Christophe: technically the numbers could be combined, but typically they are kept separate. There are a couple reasons for this, as far as I can tell - the simple answer is that it keeps a clear line between the amount of solar resource available (kWh/kWp), and the effectiveness of the system at not wasting that power (the de-rate factor). Also, banks will sometimes haircut the kWh/kWp to factor in the annual variability (creating something called "P90" which I'll discuss in a later post), and so for that they want to have the exact kWh/kWp.
Great conversation, all!