Nobody is recommending using a pulley jammer. However, we are trying to clarify what is actually happening in the system, because it may be applicable to other situations.
It would appear that the specifications of the test are such that several devices that are believed to be safe might fail to remain below the 6KN recommendation because of the way we use them compared to climbers. The chances of those factors all coming together in a real situation are remote but possible (the length of rope is so short that the climber is likely to be on a ledge rather than subject to freefall etc).
So what actually happens - the weight drops on one side of the pulley, which is unable to balance the forces equally due to the low stretch rope being firmly held on the other side by the jammer, so it acts more like a fixed curve, creating friction with the rope (sorry, the report about the force being doubled is almost certainly wrong as well). On the other side the rope is in tension, whilst the pulley / linking karabiner / jammer are in compression, which may cancel each other out in the great scheme of things, but does result in the pulley deforming due to it being forced around the inside of the karabiner. Meanwhile, much less force will be acting on the jammer (there was no damage to the rope, & presumably the cam was releaseable as it is not mentioned, but does jam in the next scenario - jammer in front of pulley) - so maybe the maths isn't as complicated as we thought...