As far as "charge the car" let's be clear that it'll simply add a (very) small amount of charge... for all intents and purposes though I'd refrain from saying it'll "charge the car" as people might misinterpret this not needing to plug it in at all.
Let's also be real about charging... I'll use a Model S as an example because it's bigger and we know the dimensions. The area from the top in sq meters is roughly 10 sq meters. at 20% efficiency this would yield 2 kWh Model 3 will be even less. Chances are that only the roof would be solar panels and not the entire surface area of the car. This puts you closer to 1-1.2 kWh
In addition, you need to leave it in the sun, this causes the interior to heat up and once it hits the max temp, the car will start to cool itself drawing power. Let's be very generous and say this leaves you netting 1 kWh.
Where I live the average is 4.21 sun hours per day, this means if I leave the car out all day in perfect sun I'd get about 4.2 kWh. This would save me about $0.50 of electricity per day.
Now realistically, I have a garage and I park in a parking garage at work. I also drive to work in the early morning when it's dark and I drive home early evening. Maybe it'd get 30 min of sun so that $0.06. If I worked 365 days a year then it would add up to $21.90 per year. If a solar roof on the car is a $2000 option then it would take 91 years for it to be worth it (in my particular situation).
For someone who only parks outside then in my area it'd take them nearly 11 years to break even if energy prices stayed the same assuming a $2000 solar roof option.
It'd be cool if it worked better or was cheaper but right now it's not all unicorns and rainbows.
That said, if I lived in a very small town and drove a small amount every day with no garage at any destination then it'd be fantastic. Might even skip getting wiring done to charge with high current at home to offset the initial solar cost.