Define “need”. If five billion people want to live as Americans live now, and Americans will all have flying cars (1), go on holiday in space stations, etc, and we want to put back the carbon of a few cubic miles of oil in the ground each year to fight climate change, what we “need” will be quite different from what we think it is now.
If solar cells become essentially free, we _could_ get there.
Also ”ground based installation is really cheap.” doesn’t matter. Putting cells in roads _may_ become even cheaper, even though the cells themselves would produce a lot less power per square meter.
”it’s an extremely difficult problem, but solving it without massive reduction in panel costs is kind of pointless.”
I didn’t claim it was easy or that it was efficient now. I just pointed out that reduction in panel costs can lead to a situation where using them in suboptimal settings can be economically viable, if installation costs are lower for that suboptimal setup.
And massive reduction in panel price has been the norm for decades. https://en.wikipedia.org/wiki/Swanson%27s_law:
”Swanson's law is the observation that the price of solar photovoltaic modules tends to drop 20 percent for every doubling of cumulative shipped volume. At present rates, costs go down 75% about every 10 years.”
(1) of course, if Americans start having flying cars, road surfaces may become a thing of the past.