Thank you for the hint, sir.
I actually thought about linearity for years until I discovered what you just said by mapping the scanner temp readings with the gauge positions, and the numbers I mentioned are accurate to some extent because I did not extrapolate or guess but by trying different resistor values in place of the sensor while watching the position of the gauge needle. Then, I used a digital contact thermometer in a small pot of water with a new AC Delco sensor to map real temperatures with sensor resistances.
Am I still doing something wrong? probably over-complicating things?