Many years ago when I ran a development laboratory, I did a statistical study on dozens of sender units and gauges to determine resistance values of sender units at various temperatures and equivalent voltage levels required for gauges.
I can’t find the resistance data for senders but I still use the voltage levels to calibrate gauges. With respect to the sender unit, I do remember that I could not substitute a resistance device of the same resistance as obtained from the sender unit and get similar results with the gauge. I suspect that with the gauges approximate 100MA current through the temp sender that there is internal sender unit heating occurring which in turns affects the actual resistance.
In your case, since you are comparing old and new units, a basic temperature/resistance test should correlate for same temperatures. The main reference I’d use is boiling water.
The following info applies to all thermal type small gauges used in the mid-sixties, whether its a fuel or temp gauge. Each gauge has tiny marks at the minimum, midpoint and maximum on the face above the numbers. The marks are either two dots or a dash line. The following voltages are required for Min, Mid and Max points. 1.89, 4.95, 8.00. A constant current/voltage power supply capable of at least 100MA is required to test. The voltage must be measured at the meter terminal.