The tube that is part of a pyrometer is designed to generate a small voltage when it see's heat... . they are very accurate... ... much more so than the gauge... . What we see is that any loose connection used will cause a reading problem... . or dirt on the connections... . we never solder these connections as a cold solder joint will give you problems... instead we clamp the connections together with a bolt.....
This technology is so simple unless you have damaged parts I wouldn't worry much...
We run an oven to cure bonding resin for friction... . this is a bell curve. . in that the we have to raise the temp, and lower it over time to get a good bond... to do this we use a digital gauge because it's accuracy is within 3 or 4 degrees... . we have 3 different guages... . and if we test the same part... . all read the same... we can interchange the thermalcouples from gauge and still get good readings.....
Gauges(mechanical meters) are usually 2-3% gauges... that means that they are accurate to 2 or 3% of full needle deflection... . lets say your gauge goes to 2000*, 2% of that is 40* so your gauge would be accurate to + or - 40*..... I'm betting the inexpensive ones we use are no more than 5% gauges... I used to work in a place that had a instrument gauge shop and we'd repair and replace high end meters... . usually a 1 or 2% meter was in the neighborhood of $200 or so...
Also remember that all meters... . no mater what they read... as long as they are electrical meters... are nothing more than voltmeters with a different face on them... . and to add to this... I'd bet that the dash on our trucks... volts, oil pressure, fuel gauge, etc are all the same meters with different faces on them..... and if the tack and speedometer are electric they are the same way... .
I agree with towpro..... this is a quick simple test... .
Hope this helps... .