I'll ask the stupid question ....
For ACW rotor systems do all tailrotor drive shafts (not the TR, the shaft) go the same way? And for the same MR direction is the effect the same in the hover regardless of the shaft direction? (Testing the drive train influence).
Having just written that, and considering the more rigid nature of the TR system, is the roll producing a gyroscopically produced TR Tq that has to be reacted the usual way and hence causing a Tq spike at the MR shaft? Being in the vertical plane and much larger in second moment inertia etc than drive shafts it would be a likely culprit for horizontal plane torque inputs when subject to roll rates.
Time to dig out the Dynamics texts, or a bottle of wine. My brain hurts already.