As someone who has looked into this in depth, vortex generators can work on a car. Want proof? Mitsubishi actually did implement them on their Evo Lancer and got a 2 to 4% reduction in drag. The problem is they needed to do a full CFD simulation and a wind tunnel to verify it.
With my old 300SD I did some CFD simulations and was able to get about a 3% reduction in drag. Would it have saved me any fuel? not really because my most of my driving was in town and short trips.
What I learned from the simulations.
1) Sizing the VGs was very important. If they were too big they just created drag. In other words don't expect VGs designed for semis to work for your car(they might but it won't be optimal).
2) Proper placement is very important.
3) To get VGs to work where you want them to you might need to address problems further up stream.
4) the number of combinations is infinite so you will never find an optimal solution.
5) Unless your car's aerodynamics is absolutely abhorrent don't expect to get more than a 5% improvement
__________________
|