Also you are correct on that the indicator is not a testor, they cannot safely build a tester that will handle the proper signal and the false signal the indicators send may not even trip all AFCI breakers even within inches of the breaker. They are a tool that will many times work but not always. The only true test of one of these is the test button on the device.
I have to disagree with part of this. I agree that for various reasons, the testers/indicators can give false indications of AFCIs not funcioning properly. However, I am not convinced that the test button on the device is a true test either.
1)If they can not safely build a tester that can handle a proper signal, how can they safely build a test button on the device that can handle a proper signal?
2)I am sure that the test button does not truely create an arc in the branch circuit being protected. It seems that the test button must necessarily be by passing some portion of the branch circuit monitoring/analysing circuitry within the device and giving an "arc detected" signal further down the flow chart.
I am of the opinion that there is no true test of the complete AFCI device. I think that:
1) If the hand held tester causes the AFCI to trip, we have a pretty good indication that the AFCI is working.
2) If neither the hand held tester nor the button on the device cause a trip, we have a pretty good indication that the AFCI is not working.
3) If the button causes a trip, but the tester does not, we have a pretty good indication that part of the AFCI circuitry is working, but I think there is still the posibility that part of the circuitry is not working. As an inspector, I do not have a basis to reject the AFCI in this case. I just have to take it on faith that the thing works.