Thanks for the comment, I understand your frustration.
Even talking about accuracy in a review like this is a catch-22: verifying consistency of testing conditions is hard, and we can’t guarantee that what you get will be calibrated the same way. Buying an NIST traceable thermometer for an accuracy test just to show that everything is within 2 degrees didn’t make sense for the scope of this review.

As stated, these were all within a few degrees of each other: https://www.yourbestdigs.com/reviews/best-indoor-outdoor-thermometer/#range
Acu-Rite gives their spec for accuracy as well as adjustment parameters for different temperature ranges: https://www.acurite.com/learn/accuracy

As other commenters have pointed out the best way to get verification is probably with a simple known-accuracy test (freezing water, brine, boiling water and a mercury thermometer) to check the readings from a non-certified model yourself.

Great points!

The one-place-decimal precision is still an interesting feature; knowing that the temperature has dropped a half-degree over the course of an hour is still notable, even if you’re not sure what the absolute temperature was at the beginning.

If you need to know when it’s freezing outside in May the +/-2 degrees accuracy is still close enough to tell you when to take action. We didn’t get a table of relative readings from our tester, but he says they were all within two degrees of each other and some of the time better than that. https://www.yourbestdigs.com/reviews/best-indoor-outdoor-thermometer/#range

If you need verified accuracy for official records, you pretty much need to buy something that comes with a traceable NIST certificate, though as you say even there the caveats are important. The price jump is so high, of course, that most of us won’t even consider it.

Holy crap! I forgot about that.

Following up on this, I see what you mean. Even a very expensive weather station like the Rainwise MkIII, while its sensor is designed to measure at -50, is only guaranteed to operate reliably to -40. Wired-probe options seem to have similar issues, even with the base station inside and the probe wire outside.

An alcohol-type thermometer outside your window seems like your best bet for economical performance on the coldest days; even mercury freezes at -40.

We definitely didn’t notice those problems in our side-by-side testing, have you already got an exchange for the LaCrosse with the defective screen?

Thanks for that note! Have you had any difficulty with setting up a return/exchange?