[aerohead]

Full-size automobiles achieve a stable coefficient of drag at around 20-mph.
Testing mods in full-scale would require at least a 20-mph test velocity.
--------------------------------------------------------------------------
At 1/24-scale,do to rules of dynamic similarity (verisimilitude),the air velocity in the test section of your wind tunnel would require a minimum 480-mph to achieve the necessary Reynolds number.This would introduce compressible flow/transonic flow effects which are not encountered in 'real' driving.
Typically,aero testing is conducted with a minimum 1/10-scale,and critical Reynolds number must be factored in.And this is 'rough' testing.
'Winners' from 1/10-scale tests would graduate to 3/8-scale,then winners from that would finally graduate to 1:1 scale testing,required to 'prove' a small structure as with a wing.
Exa Corporation's PowerFLOW CFD can analyze small features but require enormous digital CAD files,require 35-100 parallel computers running simultaneously to 'solve' a single iteration after many hours of run time.
--------------------------------------------------------------------------
You can learn some pick and shovel kinds of results at 1/24 so it can be worth playing around with it.