"btw, is it common for torsional vibration to cause significant cylinder-to-cylinder performance differences? "
No, not performance differences, that is, the imep from each cylinder is identical. What happens is that the difference in stiffness cause a different delay for the impulse to reach the measuring point and this introduces a half order (and multiples) into the FFT. It is one of the main reasons why stiffer cranks/blocks sound better. The camshaft torsionals may be the same thing, but they also twist under the valve spring load, this produces a repeatable phase error between each cylinder, and ends up causing a similar response.
However, most times 1/2 order comes from regular differences in cyl pressure diagram caused by lousy intake or exhaust manifold design, or possible temperature effects. (why does the 5th cylinder in an I6 usually knock first etc).
Yes, camshaft 2 plane balance would be at half crankshaft speed. I don't know if it is significant, but worth checking.
I've taken a week off from meetings. bliss.
Cheers
Greg Locock