We seem to be stuck on smoke and water. How about an alternate idea? Use some WD-40 or other light oil spray and spray it ahead of the area you want to test, then go drive. The oil will form streaks in the direction of airflow, or if it's turbulent there won't be streaks.
For observing airflow patterns on parts of the car that you can see while driving, such as the windows and hood, use wool tufts taped to various locations.
At least WD-40 spray oil patterns will stay around long enough to get a good look at them, although on more the steeply sloping sides there will get to be a gravity droop. If this sets in too soon, it will spoil the test. Probably more useful on local areas, one by one, for this reason. The trouble with smoke lances and other small sources to try and provide a thin stream is that all too often someone has to stand in the airflow holding and directing the smoke. No one seems to have tried a wind tunnel - what? Think about it, there are plenty of commercial die cast models, sure $10 to $40 and maybe not everyone's model is there, and 1:18 scale is probably best. Perspex top and sides and a leaf blower fan. maybe tricky to get parallel flow a smoke points, but maybe worth experimenting.
Glycol on a hot plate. A quart of fog machine fluid is $8.00 at your local party store. Drip it on a hot plate and it turns into billowing clouds of fog. It has enough cohesion to push down a pipe, details to be worked out.
The way to build a wind tunnel is to build a box several feet long with a fan at the outlet end. You don't blow air into a wind tunnel, you suck it through. That eliminates all the turbulence from the blades. On the inlet side, you build a diffuser to straighten the air flow. For a small wind tunnel, use drinking straws stacked together and/or layers of window screen mesh. The air is pulled through the diffuser by the low pressure created by the fan. The air will not be perfectly smooth (wind tunnels don't produce ideal conditions. They produce sustained, controlled, repeatable conditions). But the turbulent eddies will be very small.
The test section of the wind tunnel has to be large enough to keep the boundary layer by the walls and roof away from the test article. And you don't want the air to get constrained between the walls and the article.
It will be amazing how little wind a box fan will produce. The drag on the test article will be immeasurably small. If you've gone through the trouble to build this wind tunnel, get a gasoline engine and a propeller and make some wind happen.
Now if you really want to get swanky, make the bottom a conveyor belt that moves at the speed of the wind. This better replicates moving on the road eliminates the artificial boundary layer on the floor of the tunnel
Then you get into Reynold's Numbers and correcting wind tunnel conditions for real conditions and I'm stopping there. It's been over 20 year since I was in school and all my books are buried somewhere.
So... glycol tank, fuel pump and an injector installed in the tailpipe of your FRIEND's car... and follow behind him on the highway you say?
I remember The RoadWarrior..To understand who he was, you have to go back to another time..the world was powered by the black fuel & the desert sprouted great cities..Gone now, swept away..two mighty warrior tribes went to war & touched off a blaze which engulfed them all. Without fuel, they were nothing..thundering machines sputtered & stopped..Only those mobile enough to scavenge, brutal enough to pillage would survive. The gangs took over the highways, ready to wage war for a tank of juice