My guess is that it is a rounding issue
By definition TFR(Ref, T) is all the same for a specific material, regardless of the magnitude of of R as TFR(Ref, T)=R(Observed Temperature)/R(Reference Temperature)
Example, Iodode Titanium, close to Gr1, TFR(20°C-200°C)=1.635 (
second tab)
Means the resistance at 200°C will be 1.635 times higher than starting resistance regardless how big or low the latest might be.
That is the theory. But the practice can be different for these temperatures are average temperatures. The temperature difference along the coil can mix up things, especially in dual coil setup. Not to mention the (preferably but unfortunately not necessarily static) internal resistance of the other components of the system.
Therefore, as it is mentioned above, to calibrate/fine tune a device to a specific setup you will need proper measurements.
And for some reason I got a feeling that you will find some reliable test results here pretty soon