I've passively followed the perovskite revolution for a while now, and the constant claim is that they're cheaper. But how cheap? Nobody can ever seem to quantify it.
In a lab you can demonstrate how much material something uses, how long it lasts under given conditions, and, say, how sensitive it is to contamination. But you can't predict, you know, the late-oughties polysilicon price bubble, the ensuing long-term purchase agreements made by companies like Evergreen, and the subsequent polysilicon price collapse that sunk those companies. (I'm still not sure where modern PV cell companies source their silicon! Is it UMG?)
"How cheap?" is a question about international trade, mining, and management, not a question about materials science. Don't expect materials scientists to be able to answer it.
My personal guess is that single-junction perovskite cells will not ever overtake single-junction silicon cells for rooftop or utility scale solar. Single junction perovskite cells may be used in applications where light weight and flexibility are advantageous, like charging portable electronics, if they can be stabilized.
Perovskite cells may compete in rooftop/utility solar with conventional silicon when incorporated into tandem cell designs -- either perovskite on silicon or a stack of different perovskites with different band gaps. That gives them the potential to exceed conventional crystalline silicon module efficiency rather than merely play catch-up. The company that seems to be furthest along with this approach is Oxford PV, which is pursuing a perovskite/silicon tandem design:
The world installed about 117 gigawatts (peak) of solar PV last year [1]. If the average panel is 17% efficient that's more than 680 million square meters of panels being manufactured per year. World motor vehicle production in 2018 was 96 million [2]. To put more solar panels on vehicles than we already put in fields and on rooftops, the average vehicle would need more than 7 square meters of surface covered with panels. That's neglecting that most vehicles are still internal combustion vehicles that couldn't do much with the electricity those panels would generate.
[1] https://en.wikipedia.org/wiki/Growth_of_photovoltaics
[2] https://en.wikipedia.org/wiki/List_of_countries_by_motor_veh...
Yeah, it's the old square-cube law in action.
That is a bit below what a very aerodynamic car needs to maintain highway speed so sort of relevant from a range extension perspective but 100% efficient panels and 5m^2 being fully illuminated at once are not very realistic assumptions.
I could see electric trucks(non commuting) + tiny home on wheel with big arrays potentially being viable though. That + starlink would make a pretty cool working remote combo.
Putting PV panels on a vehicle can give you a few miles per day - enough to get you to a plug if battery drains too low. More an emergency tool than viable power source.
Rooftop mounted, optimally oriented PV systems at Seattle latitude have an annual capacity factor of ~ 14%. That means a 1 kW system will average 140 W output over a year. That system has a module area of about 4-7 m^2, equivalent to the area available on a sedan.
On a car that has suboptimal module orientation and solar exposure, "power density" (RE: Smil) is really low.
A way to think of it is to imagine a gas-powered car with a gas tank that magically refilled at a rate of one gallon per week. That would be a great feature, but it wouldn't really change what you do with the car. It would mostly just mean spending less on gas over the lifetime of the car.
In the end, a car costs a lot of money and provides a couple square meters of usable space. Roofs of houses are a lot bigger and can supply the needs of a household, but if you're in the business of energy production and want to generate energy on the scale of a major power plant, it makes far more sense to use cheap land in sunny places away from cities.
"new solar cells are 94% of the price than existing mono-crystalline cells"