[tpepmeie]

I have concluded that GM understood the MAT sensor behavior in the factory location and compensated for it in the calibration. The table is "F26 INVMAT MAT FACTOR VS. MAT". This modifies the air charge density based on the MAT value. What's interesting is that the OE calibration doesn't match the true physical air density calculation. It has approximately a 10% offset.

In the attached graph, the blue curve is the mathematical air density curve. The Orange curve is the OE calibration. Horizontal axis is air temp. Aside from some obvious high air temperature protection built-in (the dip after 56C), you can see the OE curve is approx. 10% under the real curve. The grey curve is the physical curve minus 10%. Fits remarkably well with the OE curve.

Edit: the OE cal richens the mixture, which make sense if the true air temp is cooler than being reported by the MAT sensor.