It was purchased because the sensor it uses is theoretically a fuel cell sensor, that is, a type that guarantees reliable measurements.
Unfortunately, it turned out that it does not have the sensitivity required to measure accurately and reliably in the range around 0.25 to 0.30, so that someone can rely on it and be sure that if tested on the road, they will not have exceeded the limit.
Essentially, what it does is capture some peak values roughly close to the time alcohol was consumed, without being able to detect in detail and with repeatability the decrease over time.
Each person metabolizes alcohol differently, and this also varies each time depending on conditions, such as how much time has passed, if there was food, etc. So if you cannot know for sure at this moment what the exact alcohol value would be that a device would show in a roadside test, there is no point in using it. Fines nowadays are unforgiving.
Below are the measurements I made, proving that it works but with very low sensitivity and poor repeatability.

I drank 128 ml of wine within half an hour along with food and finished at 23:15.
I drank 110 ml of water and then took double measurements every half hour with a 2-3 minute interval:
23:30. 0.00 0.00
23:45. 0.00 0.00
24:00. 0.00 0.00
24:30. 0.00 0.00
01:00. 0.00 0.00
With the wine, it measured nothing.
At 01:05, I drank 27 ml of whisky and it measured HOT, meaning it exceeded the scale limit it can measure. This was logical because I measured immediately after drinking, so it measured vapors from the mouth and not from exhalation.
01:30. 0.38 0.19 0.14 three measurements with a 2-3 minute difference between them with three different values
At 01:35 I drank the remaining 27ml of whisky.
It detected nothing, despite the previous history.
02:00. 0.09 0.00
02:30. 0.00 0.00
It is certain that the rest of the model series uses the same sensor, so there is no need to pay more money for something more expensive, i.e., better. The others work the same way.