"Horridly inefficient" relative to what?? The intent here is to take advantage of the historical, and continuing decline in the per watt cost of the cells. At the same time, the long exposure strength of the cells would aide the gen3's mpg at the margins (perhaps a 3 mpg improvement per tankful)! If the cells only produced enough juice to run the AC on a humid hot 98 degree Dallas Texas afternoon, it may be a worthwhile option, if the price wasn't too high.
Current on the market peak efficiency for a residential solar panel is ~25%. So 75% inefficient, pretty horrid, and that's with ideal placement.
Photovoltaic cells are only at peak efficiency for ~8-16 hours a day, and that's with ideal placement, no clouds, etc... Ideal placement is never on a horizontal plane, so the roof, hood etc of a automobile, are naturally terrible places to place a photovoltaic cell. Not to mention, you'd always have to park in direct sunlight (especially your scenario in Dallas, you would potentially be damaging the HV battery pack, charging in such heat extremes.)
Next we have to look at the fact that our systems are high voltage, not 110v, so the number of cells required wired in series etc... Plus we have to look at the electronics systems that collect and regulate that energy captured by the cells, which also aren't 100% efficient.
Using the Hyundai option as a baseline, if you got the claimed 800 miles a year/ and drove 12,000 miles a year that's an efficiency of .066 % if my math is good. For this system to even be able to give the Insight 2 miles a day of EV driving, the battery pack size would have to be increased. Not to mention this is an option that would take years to pay off. (example below, of how fiscally irresponsible this is currently)
Even at $1000, that option would take (assuming 12,000 miles driven/year at 50mpg: $3.00 a gallon of gas, and that the photovoltaic cells hit Hyundai's claim of 800 miles a year)
720$ year w/o vs $672 with a year with = 48$ a year in savings.= 22 years to pay for itself.
And that's with near ideal numbers and assuming that in the 22 years of owning the car, you don't have any damage to a cell that required repair. (Like hail, or debris while driving, and that the cells are still operating at peak efficiency 22 years later)
Just to clarify, I don't have an issues with Solar technology, and worked in the alternative energy field for 5 years. When it comes to automotive implementation of technology, I'm all for it, at the current state though, this technology is a glorified gimmick, and is years away from any meaningful application.