There is at least one ground effects car out on the salt I can think of.
Ground effects is a good way to create downforce (negative lift) but it is subject to some limitations and here are a coupe of things to bear in mind:
- Increased Pitch Sensitivity
- Increased Yaw Sensitivity (in relation to downforce created at larger cross wind angles)
- Limits Suspension Geometry
- Can impact upper body flow
Unless you have access to a rolling road wind tunnel (I wouldn't trust CFD for absolute lift values - especially not steady state) I would tread very carefully with developing grounds effects. Make speed increments slowly, check live downforce data (easiest way is tracking suspension loading at speed) for signs of the ground effects "stalling" (not really the right term - but it makes the point).
Teams with big budgets have backflipped cars into tree lines when getting ground effects wrong (cough... Mercedes).
You can build a scale model to test the effects.... however, you will need to flow much faster than the full scale car to get in the correct reynolds regime. Once at the right reynolds number you can essentially scale the force values. Then you will need to consider the design of the scale tunnel you are constructing carefully, incorrect boundary layer (i.e. anything above a 1mm displacement thickness) will give false results.
This is a very tricky subject, F1 teams invest billions on this question alone.
Don't want to discourage you, just trying to keep you safe. Please try it, but increment the design of the ground effects slowly, use the great white dyno and read the data carefully.