I'm a graduate student in a research lab that does research on solid state sensors for volatile organic compounds, which is essentially what breathalyzers are. I also do the programming of the instruments acquiring data (in Labview). I would find it pretty unlikely that the source code itself is especially hard to write, unless there is something unusual going on. Either the company is just stonewalling because they can, or the code either: A. Reveals some proprietary method of detection through the calculations performed or B. The corrections that must be gone through are so involved that they are actually worried that their sensor may be invalidated.
I wouldn't be surprised at all if it was B, as there is a reason that breathalyzers have to be recalibrated relatively frequently. It is pretty difficult to make a sensor that doesn't degrade over time, and correcting for this degradation can be quite complicated.

Although this repetition you speak of is important, it is far less beneficial than exponential. (0.993)^3 is actually less accurate (0.979), I would guess you meant to do (1-(1-0.993)^3) = 0.9999997. Also, re-measuring data does not give an exponential error decrease. Most signal to noise levels go up as the square root of the number of samples taken. This is assuming the error isn't systematic.

Although funny, this type of thinking does occur in research. With the onset of the internet and search engines like SciFinder, it is quite easy to have what you think is a great idea and then find that someone else has published a paper with that idea earlier that year, or even just a few months ago.

People in my research group regularly are frustrated with the feeling that anything that is easy has been done, and all the good ideas are taken.

The problem is that when you see someone else has already started a certain line of research, it can prevent you from pursuing that line of research or one that is similar. My adviser even recommends only taking a cursory look at the literature before pursuing an idea, because otherwise this exact effect can discourage and frustrate research.

Working in a group can be especially beneficial then, because if one person has a breakthrough in a certain section of science(chemistry in my case) then the members of the group who are frustrated can either assist, or more likely, spin off and do something tangential that also hasn't been explored.