Ok, that was my mistake then. I didn't know where the Porsche track was, but knew Edmunds is in LA. When you said Carson, I was thinking it was tested in Nevada, but that's Carson City, so I was figuring the A4 and BMW were tested in LA and the Model 3 in NV for some reason. I would say the pavement being the same piece is the most important for getting consistency across different cars.
Yes, the inconsistency is the main problem with CR's findings. Hopefully Elon is right and that it's an ABS calibration algorithm issue that can be fixed OTA.