The measurement more so than the apparatus would be so humidity & temperature dependent that it very well may be a more accurate relative humidity or dew point measurement device than it would be for alcohol content of beverages.
The beverage data sure does look good though, varation perhaps due mostly to slight changes in ambient conditions, as well as matrix effects like sugar content of lower proof beverages.
Professionally, directly measuring Alcohol By Volume (ABV) which is the property that is regulated, could not be most accurately accomplished with any technique that comes to mind.
In the lab, alcohol content is always best determined by weight, then converted to volume percent using a number of traditional conversions.
You might have something there with a fuel cell approach, but if you are detecting mass or relative mass the only way to arrive at volume is by conversion.
Even when you calibrate using precisely prepared volumetric reference materials, if you are still detecting mass then you would be better off using precisely prepared gravimetric reference materials instead.
Gravimetrically I can be traceable to NIST using only traceable reference weights without requiring NIST ampoules of reference alcohol.
The measurement more so than the apparatus would be so humidity & temperature dependent that it very well may be a more accurate relative humidity or dew point measurement device than it would be for alcohol content of beverages.
The beverage data sure does look good though, varation perhaps due mostly to slight changes in ambient conditions, as well as matrix effects like sugar content of lower proof beverages.