So, what is the conclusion? But what about the absolute accuracy? I think I would trust in the very first measurements. If the instrument doesn't work with temperature compensation than it should be accurate at the standard room temperature. Isn't it? So, a "warm-up" can produce much more stable but drifted results.
Absolute accuracy is difficult to assure without a calibration reference. Such references are expensive, and themselves need regular calibration against standards. This is not within the scope of entry level instruments or users. There is no particular reason to assume that a cold reading is going to be more accurate than a warm reading. The conditions under which the instrument emission mode was factory calibrated is not explicitly stated, although one could assume that they are the same as reflective measurement :- 23 degrees C. Graeme Gill.