Thats not actually a win for the company who does the car advertising.
Its very shot-gunney. And very pricey for a shot-gun approach. ($700 per car at 1,000 cars = $700,000 per month for an ad approach which is pretty much random.)
The only companies I see this working for are those that provide products like cable - or some other such utility that everyone requires.
And if they saturate a small area. Ads on 1000 cars all over USA is probably useless. But 1,000 cars in one mid sized city may get some good word-of-mouth going.
Even after that - buying ads on the side of the buses is a much better idea. You know the exact route and can concentrate your efforts in a given area.
Case Study:
In Bombay - buses are red in colour. A few years back when internet was still on dialup - a company came up with an offer: install their dialup browser that shows banner ads and call up a local number to browse for free.
The company had their eyes set on orange (as their mascot was a tiger). They paid money to get all the buses of the city painted orange for a couple of months - with their name and offer printed on the sides.
Became very popular very quickly.
And got a lot of free publicity too - as all the local media - newspapers, tv etc gave their opinion as to such advertising should be allowed on public buses or not...
Case Study # 2:
If you're doing car advertising - spend a bit more - and instead of going with decals - go with something that stands out strikingly in traffic:
Eg:
http://commons.wikimedia.org/wiki/File:Red_Bull_car.jpg
Phil may have a bit more to talk about this... how to make car advertising work...