As there seems to an emphasis on temperature and when to take it, this is equally true for when you measure/set RPM.
Many times convenience is one's own worst enemy whether you use one method or both. Having an engine on the bench where everything is handy and out of the way usually means that many people will do an initial setting and think "I'll check under/after a hot load later, this is good enough for now".
This is true for many people (maintenance or rebuild), not all, there are some that have the helpers or discipline to do it right.
Back to tachs; what is true for IR temperature measurements is true for properly reading/setting rpm limits. You get different results cold, warm, hot, hot after load, post brake-in, elevation, etc.
Both methods require you to measure at elevation, post break-in, during and after load. The reason being is thermal runaway, which also shows up on the tach as excessive rpm that does not manifest at cooler temperatures. Depending on tool resolution you will notice this if it is in range of your tool.
That said, a high resolution IR thermometer is much cheaper than a high resolution tachometer (some cheap tachs are as bad as +/- 1000 rpm; that means it could be as much as 2000 rpm out!) but both are very useful tools.