Keep in mind that the temperature probe for the MES (at least my gen 1 MES 40) is located right above the heating element. So it "sees" a higher temperature than a probe placed elsewhere in the box will see.
Without a fan to stir the air in the smoker, you will see large differences depending on where a thermometer or probe is positioned.
Years ago, when I mounted a small thermocouple directly to my MES's temperature probe, I found that there were still errors in the MES readout. But they weren't as great as what I observed when I measured temperatures with a sensor placed far from the MES's sensor.
This is especially true when you have a big piece of moist, cold meat in the chamber. At that time, I speculated that there may well be a method to
Masterbuilt's madness.
Despite the lack of apparent accuracy, I still have not modified my MES's temperature control system, and I get good results, with cooking times to meat internal temperatures that match well with other people's on the forum.
I do, however, use an Amazin pellet maze for my smoke source. The chip burner of the original MES never did give me consistent or adequate smoke.
The meat probe of my MES tested to be surprisingly accurate, and I always use it with confidence.
I tested the MES meat probe with the dry block temperature calibrator that I built years ago to calibrate laboratory thermometers and data acquisition system probes. This relies on using a NIST certified reference thermometer.
I tested the MES's chamber sensor with a thermocouple and readout that I had calibrated in that temperature calibrator.
Testing temperature probes and readout systems should be done carefully, realizing that just having them in the same oven or smoker is not enough to keep them at the same temperature. We routinely observe differences of 15 degrees C (27 degrees F) between locations only a few inches apart in expensive laboratory ovens.
Unless an oven or incubator is vigorously stirred, that's to be expected. And even well-stirred ovens will show surprising variations from place to place.
Also consider the speeds of response between different probes and displays.