I follow your argument, I just don't think it matches reality. In any system analysis, you have to consider all the relevant factors. If you wave away importan

I follow your argument, I just don't think it matches reality. In any system analysis, you have to consider all the relevant factors. If you wave away important factors for the sake of argument, you get the wrong answers, and end up in long arguments on internet forums, repeating yourself a lot. Let's game out some scenarios using your numbers. The choices are between a $1 array that gives 100 watt-hours, ($0.01/Wh) or a $50 array that gives 1000 watt-hours. ($0.05/Wh) Rooftop solar, grid power is $0.001/Wh: You don't buy either array, since neither is cheaper than grid. This is true for most of the industrialized world. Rooftop solar, grid power is $0.025/Wh: You buy the 100 watt-hour array, since the power it produces is cheaper than grid power. You don't care about total wattage, since you can get all the watts you want from the grid. This is true of places where either electricity is expensive, or PV is subsidized. (Hawaii, Germany, etc) Rooftop solar, grid power is $0.1/Wh: You do buy the 1000 watt-hour array, since it saves you from having to buy expensive grid watt-hours. There is no place on Earth where grid power is more expensive than both single-crystal and multijunction PV. Rooftop solar, offgrid: Here you do strongly care about array output... but you're only offgrid in very rural locations, where land is cheap. In practice, off grid solar is never limited to just rooftops! If you need more watt-hours, you build bigger arrays, rather than paying five times as much per panel. Space: You need every milliwatthour, and each gram of spacecraft costs hundreds of dollars anyway, so multijunction PV suddenly becomes price competitive.