OK we seem to be the DIY kings, so what would it take to build a wind tunnel? Can a body shop spray booth be used? In which case what kind of air speeds are necessary to obtain results? I am sure we can build the contraption to blow out smoke in streams ...
You mean for full size? Whoa.
Now if you are talking about 1/4 scale models or smaller, then ok.
Here is data on the MIT wind tunnel.
http://web.mit.edu/aeroastro/labs/wbwt/index.html
One of the links says that it uses a 2000 HP motor. The good news is that for airspeeds of 15 mph you only need 350 HP, but I suspect I am not reading that right.
If I were going to instument and test small models, I would start out and make a platform that stuck out in front of a car in the clean airstream, have the model on it, and drive the real car around and take data points.
(air temp, air pressure, velocity, drag/lift loads on model, for a start).
measure the loads using strain guages.
I suspect the biggest problem is that in real life, a car is moving on a stationary road surface thru stationary air. All the wind tunnel models I know of a car have the "road" not moving, AND the car not moving, which is different than what really goes on. So the airflow under the car is different in the wind tunnel vs on the road.
To make that less confusing: In real life, the air does not move relative to the road. In the wind tunnel, it does.
When you model aircraft, the "road not moving" problem does not exist.