Shockme, I appreciate your desire to do as good of a job as you can.

The statement "Anything, even if..." I criticise as it is, at best, sloppy logic. When that particular approach is used, often only one element is looked at, and not things in context. Often, it is also mere supposition, and there is no data to support it.

Our recent history is full of actual examples of such good intentions gone astray.

Just as an example - and I do not want to get into the entire 'gun control' debate, but just want to ilustrate the point - one often hears the 'just one life' argument used in support of additional gun laws. Yet, this argument completely ignores:
- Lives saved by firearms;
- Changes in criminal behaviour to target unarmed victims; and,
- The failure of experience to validate earlier measures of a similar type.

With AFCI's, I'm afraid that we're making many of the same errors.
The lack of AFCI devices precludes their use to protect parts of circuits, or existing circuits. This is ironic, as it would seem that older wiring would be more in need of this protection than new wiring.
The continued failure of AFCI makers to deliver as promised undermines their credibility.
The '96 code panel mandating a change in th '99 code undermined the claims the NFPA makes as to 'open' or 'transparent' code development. That today's vote can be binding on tomorrows' vote violates the very purpose of voting.
The lack of any acceptable 'tester' makes you wonder if that test button actually does anything.

Most important, the "but one life" argument ignores the unintended consequences of attempts to circumvent the code. How many AFCI's will remain installed a year after final inspection is one issue. You can be sure that there will be casualties as the untrained try to 'fix' things.
The substantial added expense will also serve to steer folks away from responsible electricians.

So, the jury will be out for some time. Unfortunately, it is far easier to pass a rule, than to repeal it.