Let’s just assume that a 1 square meter solar panel will produce about 1 kilowatt-hour each day. That’s a generous assumption, as the 1 kwh/day figure comes from fixed solar panel installations, and some cars will be parked under trees, in garages, etc, and the solar panel will not be perfectly aligned with the sun. In any case though, we’ll go with it. That corresponds to an average power of about 40 W, spread over the entire day.
We’ll also assume that US auto sales total about 1 million per month, which is approximately correct (it used to be more than a million, has recently dropped to less than a million). And for the first few months/years, we’ll pretend that everyone who is buying a new car is replacing a non-solar-equipped car.
So, in the first month, the total energy production from all solar-equipped automobiles will total 40 MW. In the second month, 80 MW. After the first year, 480 MW.
If all automobiles in the US (about 250M) were so equipped, as might happen after 10 years or so, their total energy production would be about 10 GW. Of course, this number is too high, as the solar panels on older vehicles will become damaged/dirty over time.
For comparison, a medium-size power plant produces about 1 GW. Additionally, the total power consumed (on average) by all vehicles in the US is about 600 GW (based on 400 millions of gallons used per day and 0.13 gigajoules / gallon).
A single 1-square-meter solar panel therefore appears approximately 1/60th too small to power the average vehicle, even given all the optimistic assumptions above, and assuming 100% battery efficiency. This assumed a solar panel efficiency of 15% or so, so even a perfect 100% efficient solar panel (impossible) would still be about 10 times too small.
Solar sucks.