An excellent question!
No, I have not compared this to a certified gauge or pressure source for absolute reading accuracy - that is within 0.5 psi as I recall from the manuals and specs. For absolute accuracy, I'd love to get this one:
http://performancetuner.net/DTPG.html - it has 0.03 psi guaranteed absolute accuracy and resolution to 0.01 psi, but this is major league overkill!
However, from a relative readings perspective, with the Accutire gauge, I can set my tires to within 0.1 to 0.2 psi psi of each other, since the readings are totally repeatable from one to the next reading - the slight drop each read is due to the very minute release of air each time I press the valve in.
And that is what I was looking for with this gauge! I set my tires to a 36 psi reading and adjust up or down based on tread wear (which a cheap depth gauge can easily measure - the penny or quarter test is not sufficiently accurate really).
Plus, I can watch my tire pressure drop over time (as is very normal) and they tend to stay within 0.2 to 0.3 psi of each other as they lose pressure. If they differ by more than that over a 2 to 3 week period (as happened to my wife's tires a few weeks ago), I know that some tire is leaking more than normal.
In my wife's car case, it turned out to be a small screw that had embedded into the tire! After repair, I am back to the same relative change weekly.
By the way, with ultra low profile tires, like the 30, 35 and 40 ratio tires (and the 45 ratio I prefer to use for more sidewall and comfort) people are doing here, it is vital to check the pressure often - ideally ever week to two weeks.
Because there is less air in the tire overall compared to a high sidewall tire, it does not take much air loss (normal over time) or temperature changes for the tire pressure to change by more than you might expect.
Z