Panels must be positioned to face the sun, must be unobscured, must be in daylight during prime hours to make it worthwhile. Car roofs are often none of those things.
You want car parks turned into solar stations? Then mount them in the car park over the cars, angled at the sun permanently. Its a simple change, cheaper, and works many many times better. And gets crap off your car making it lighter and reducing drag.
This is an under-appreciated point by the solar car advocates.
In order to get anywhere close to a practical self-charging car, you have to put panels on an area equivalent to covering the entire surface of the car (including the windows!).
For a conventionally-shaped car, approximately 1/3rd to 2/3rds of those panels (depending on the orientation of the car w.r.t. the sun) are producing essentially zero electricity. Either because they are actually facing away from the sun, or just at such an extreme angle to the sun.
In other words, you get between 50% to 200% more power by taking those panels off the car and slapping them on a roof. The difference in practice will be much greater in urban environments because buildings tend to shade the street (where cars park).
You don't think the same would happen if wide spread adoption of solar started happening?
Arguing: "Technology isn't good enough yet, so we shouldn't try"... how do you expect to make it good enough?
A car having anything remotely close to useful solar charging requires so many goldilocks conditions to be just right that it's not worth the complexity, and it's better to invest equivalent resources into more practical solutions.
Just to make my point really clear, This whole thread is generally arguing: Don't do innovative things, because there are other known (non-innovative) things you could do.
Cars that charge themselves in the sun is a no brainer, is it going to replace needing to ever charge them, not yet... But if we don't try we won't improve that technology.
The ICE brought about big gains for its use case even as a prototype; ICE cars don't produce poop and they wear out a lot less quickly than a horse. In comparison, flying cars or amphibious cars never really made it off the ground in a big way, because the "innovation" there resulted in a hybrid machine that did two jobs poorly and impractically, rather than having two specialized machines.
If it is really easy to point out inherent problems, like, say, the fact that most of the surface area on a car is not generally oriented at an angle to the sun conducive for generating electricity, and those aren't problems that would be solved by "innovation", why try to squeeze water from a stone?
The same thing won't happen with solar, it needs active use in novel areas. Cars are out in the sun, why not collect it? Panels aren't exactly a super rare resource, we could and should make piles of these and stick them to everything, it's literally free electricity and basically 0 maintenance.
Are they? Cars spend a fair amount of time under things that would block sunlight; there are garages in all sorts of buildings from your bare-bones detached house garage to multistory affairs, and even if they're outside they're often parked under trees, or in streets and lots shaded by buildings. So it makes already intermittent solar even more situational.
In fact, in hot climates, people take care to park their cars in shade, because the last thing people want is for their car to be baking in the sun.
Or the other alternative, the shipping of petroleum on huge trucks to millions of locations.
But yeah, I'd draw the line at solar panels on car roof tops too... the excess is too much.