Just thought I’d add my few pennies worth.
A little before Christmas I fitted a digital analogue look-alike meter and interface I had designed to monitor the coolant temperature as the ECU sees it. This meter is a 17 segment LCD that looks like an analogue meter and the interface connects in parallel to the 2 terminal temperature sensor above the starter motor feeding the ECU.
As some of you know I had been monitoring the ECU coolant temperature sensor with multi-meters for some time and had got a reasonable understanding of the variations of temperature. However, the sensor is non-linear and I had only the workshop manual to give me an estimate of top end calibration based on when the radiator fans come on. I could only get a measurement for when the low speed fans come on, I really wanted a measurement when the fans went to high speed but I could never get them to come on – and I really tried hard.
So I gave up and designed and calibrated my interface according to my ‘best guess’ from the data I had up to that time. I set the meter to read (estimated) from 86C to 110C in 1.5C steps. This being from just above the stat opening temperature to just above when the fan goes to high speed – panic level.
When fitted and run for a few days I was rather chuffed

to see that the temperature sat at 50%, mid-way, straight vertical position for most of the time (estimated 98C), driving on the level no matter what speed. It drifted up and down according to the hills and traffic conditions. The low speed fans coming on above 2 segments above 50% and off again below 2 segments below 50% ~ on = 101C off = 95C. The hottest I got it to was 4 segments above 50% estimated at 104C at which point the Mason gauge was showing about 90% of full scale and the alarm was just beginning to sound.
I should point out (although obvious to most I suppose) that this is the temperature about 2/3 rds up the block and will be lower than the temperature measured at the top front end of the head which is where the dash gauge sensor is.

My theory: -
So if the stat 'normally' runs mid-way at 90C, the ECU sensor is at 98C the outlet temperature to the bypass and radiator should be up at about 102C so you need about 20C of cooling via the radiator (and pipes) to maintain the stat at 90C assuming that the bypass and radiator flow at the point of mixing is 50/50. This accords with my measurements of hose temperautres either side of the radiator.
I was going to try recording the temperature along side the thermocouple reading from the stat housing on the Maplin trip recorder sat above my left shoulder to also catch the rev counter in view.

BUT as some of you may be aware my Bongo was written off a few weeks ago and my meter and interface are now in a pile of bits in a bag.
When I get a new Bongo I’ll be re-fitting it and starting again.
Regards Grahame