I'm guessing finding a treadmill capable of running at 200mph is quite tricky.
Although, what about a rolling road, the sort of thing they use for dyno testing, or does the fact that the whole surface underneath the car isn't moving relative to it mess that up?
Further to this question earlier.
Yes, the belt needs to be flat and the boundary layer on it made perfect before it gets to the car. All 4 wheels need to be rotating. The flow field around stationary wheels is quite different from that around rotating wheels, so measuring a car with the wheels stationary is doing so in a non-representative flow field and is therefore not only wrong but not correctable by applying an offset or correction factor.
It means most road car aero has been done wrong (in terms of drag, lift and radiator flow) for decades.
We did a comparison for a road car client in the early 80s. They had a prototype car and a 1:4 model.
They had wool tufts fitted and had data of photos of the real car, the real car in a conventional full scale tunnel with stationary wheels sitting on the force scale pads and the model on a 1:4 tunnel.
We made model rotating wheels and tested the model in our 1:4 tunnel with moving ground.
The only correlation with the real car was our model tests.
I forget which way round it was now, but on one test the radiator air went through the radiator and exited through the front wheel-arch in the other it exited under the car, hence (one of) the inability(s) to correct the error.
A fixed road tunnel is OK for wind noise from mirrors and dirt build up on window type work but worse than useless for forces because people think it is right when it is not.
As ever having no data is better than having bad data IME.