You are working under the assumption that the "legal" setting performs legally (1 pull 1 shot). This is contrary to my findings. As far as sharing my results, I have already stated that I won't and why. I didn't have altruistic motives when I did the testing. Myself, and the gentleman who arranged the markers to be tested saw it as an opportunity to make money. While I could probably sell a dozen or so "idiot" trigger systems to people here, if frames with microprocessors were deemed against the rules, we could make some actual money.
It will likely never happen for scads of reasons (several of them outlined here). You don't take the threat of litigation seriously, I have a business, a wife, and children, I cannot afford to take those kinds of chances.
As for my equipment and methods, sure:
http://arduino.cc/
http://processing.org/
A few alligator clips and a USB cable.
I set up a passive switch listener for both the trigger switch and solenoid. Was surprised that this proved adequate for almost every setup. I programmed the Arduino to report over virtual serial (USB) to my processing app which formatted the data into readable console lines. The first version worked a bit differently but the version we ultimately used printed the data like this:
1 Second starting at (startmillis) and ending at (endmillis)
x switch / x noid
That's it. When the RoF increased we started to see the number on the right consistently exceeding the number on the left. I can't speak to the settings because none of the markers tested were mine, nor did I alter the settings in any way. We only asked that those who would let us test their guns set them to "tourney legal semi". Three of the guns tested belonged to the gentleman I mentioned earlier and there is absolutely no way he does not know how to set his equipment properly.