You are (in your own words) making the assumption that the temperature gauge fitted is correct and the gauge in question is fitted to, and measuring the temperature in, the top hose.

My point is that, although you may well get 140ish ohms when the fan cuts in, you have no idea what the actual temperature is at the sensor in order to make a comparison. Yes the fitted gauge may say 109, but the temperature could be, and is highly likely to be, much higher in the water jacket itself? For example the sensor might be sending a reading of 140 ohms to the ECU so the fan is switched on, but you don't know whether it is the right temperature. In other words, the range of resistances produced by the sensor may well be within spec on the multimeter, but you won't be able to confirm whether the correct temperature range corresponds to those resistances when relying on the fitted gauge in the top hose. The crucial thing to test is whether or not you're getting the correct resistance at the correct temperature. Just checking the resistance is not enough to ascertain wether the sensor is good or bad in itself, unless you can cross-reference it with temperature.

As you say, the only reference temperature you can be 100% sure of is the ambient. The rest is based on assumption.

__________________