There is no reason both devices should be accurate. Both of them could be wrong. Still, a 2 deg C difference does indicate one of them is off.
Proper calibration is not an easy task.
Calibrating for 0 deg and 100 deg. C is straightforward. For every 500 meters in elevation the boiling point decreases by ~0.5 C. The weather affects atmospheric pressure as well. Freezing points are also affected by pressure. But you need to calibrate for a different temperature range.
Calibration using other liquids, or mixtures of liquids, with know boiling points within the range you need could be hazardous without access to a laboratory with a chemistry hood.
sells calibration tools. They offer safe calibration liquid mixtures for $99/liter. You would also need a way to heat the bath. Their certified calibration simulators for digital thermometers start at $159.
ThermoWorks also sells high-quality, digital, immersion thermometers for brewing. These are NSF (National Sanitation Foundation) certified. I would just use one of their immersion thermometers.
Another option would be to use one of the thermometers you own. I'd assume the error in the temperature range of interest is constant. I would run a series of tests over a small temperature range. If you see a difference in quality I'd just use the optimum indicated temperature from there on. I doubt atmospheric pressure changes would be affect film/transparency development.
1/ I am just a happy customer who uses their products for cooking.