Ok, something very interesting that I was thinking about and it shows something that I was not really aware of how much of a factor it would be... and this is that it takes a lot more air over a longer period of time and at a more consistent rate so that the reed has time to stabilize, and give you more accurate reading.
I am no expert, but from what I see, the meter has not had enough time to decide exactly what it is that it's hearing to stop moving between too sharp or too flat and therefore, hard to say "yes it is exactly 2 cents sharp" or whatever, and deal with it.
I think I would want a goal to have a tuner that can give me a minimum of 6-7 seconds of steady tone from the deepest of notes (the ones that take the most air to sound properly), and from what I see on my own accordion, playing the deepest bass note requires at least a 1/2 to 3/4 amount of travel from a set of full sized bellows. Now, that or (obviously) a motorized suction or push of air at a constant velocity from a (either) a very quiet and close source or a larger long hose to keep the local ambience as quiet as possible and not "infect" what the computer or device is hearing.
Fun to speculate.