That is a potential problem, but I don't see it as a serious one. These things reach equilibrium pretty slowly, over a couple of hours, or so. So, unless there is a steady stream of CO2 mist bubbles gathering under the "horn" of the device, then dumping their remaining CO2 into the air gap, all that should happen is that the indicator will "overshoot" the true CO2 value, but settle down to an accurate reading eventually. If the ppm of CO2 in the bulb of the device is above that in the water, the CO2 will migrate back to the tank water.
I think the problem can best be avoided by choosing a good location to mount the device - where there is good water circulation, and at the opposite end of the tank from any mist generation. Those CO2 micro bubbles can't exist for very long in the water, before the CO2 diffuses into the water, leaving a microbubble of other gases. So, I suspect that most of the bubbles are "inert" by the time they are swirling around the opposite end of the tank.