Where do they claim that? Do they say which kind of sensor they're using? Electrochemical? Metal-oxide? These sensors usually require advanced periodic calibrations, you need to correct for temperature and humidity zero and sensitivity dependence, model sensitivity drift as the sensor ages, take into account cross sensitivities to other gases.

I'm not disputing anything, I'm sure they do all that. I'd love to read some detail about it or a third party review that did a comparison with a reference device. It would also be cool to place some of their detectors side by side and see how much they agree on the measurements.