Arc-fault breakers are a RED-HOT topic of contention among electricians, inspectors, building departments, code-making persons, etc.
Here's my personal opinion. They're one step worse than worthless. Here's why.
When they first became required, I was a bit skeptical of their ability to 'detect the characteristics of an electrical arc'. The only thing I knew of that can actually detect an electrical arc is a line protection relay that the power company uses on their transmission and distribution lines. Not only can it detect an arc, it can tell you how far away the arc occurred. It costs about $3500.
I decided to conduct an experiment. I hooked up an arc-fault breaker in a panel, and brought the two wires out to a test bench. I wired a cord cap on the end of the wires. Next, I cut the end off of a cheap two wire household extension cord. I then stripped about 4" of insulation off the end of the cord, so that all the fine strands of wire splayed out. Next, I plugged the cord into the cord cap that was connected to the arc-fault breaker.
Lots or arcs, lots of sparks, but the breaker didn't trip. I shook the end of the cord around. Even more arcs and sparks, but still no trip. Next, I shoved the bare ends of the cord onto a steel table top. That tripped the breaker. (Don't try this at home kids, I had insulating gloves on as well as other protective gear, plus 19 years of electrical experience.)
My conclusion is that the arc-fault breaker failed to detect the characteristics of an actual electrical arc that certainly would have caused a fire, had it been near anything combustible. The fact that it tripped when directly shorted seemed to indicate that it wasn't defective.
The reason my opinion is that they're less than worthless is based on the false sense of security one might have if these are installed. Frayed cord...no problem...I've got arc-fault breakers. Impossible to start anything on fire.
They might be improved by now, but when they first came out they were nothing more than a breaker with a magnetic trip unit set at about 75 amps (as opposed to about 150 for a normal breaker), and a 30 milliamp ground fault trip unit. One of the biggest complaints I hear about them is a basic vacuum cleaner will trip them. The starting surge of a 12 amp vacuum is about 100 amps.
All code-making committees have at least one manufacturers representative on them. The original purpose was to be sure that something they were thinking of requiring could in fact, be made. In the case of arc-fault breakers, it was the manufacturers that pushed the requirement the hardest. Hmmm......