Re the asymmetry in the arms. I'm still not sure what's going on here, but will add a few points. The problem is in the rig, the not the obj. The chest seems symmetric, with the asymmetry starting in the collars. Setting the 'orientation' to zero for all the arm actors seems to cure the problem.

I'm not advocating that you do set the orientations to zero, just noting the fact that this seems to cure the symmetry problem.

Now here is where I get confused. Starting with the orientations for the arms zeroed, I use the Joint Editor to set the orientation of the rCollar to "orientation 0 0 -8.5". -8.5 is the Z orientation, as the three values run x, y, z respectively. The result is that a vertex I am monitoring on the collar moves up.

Next I change to the the other collar, the lCollar, I set this to "orientation 0 0 8.5" (the inverse of the rCollar). The result is that a vertex I am monitoring on the collar moves down. Hang on! Shouldn't it move up? Irrespective of whether I set the value to 8.5 or -8.5 the lCollar moves down (and vice versa for the rCollar). This seems to be where the asymmetry in the arms is coming in. Setting a value for the xOrientation introduces an asymmetry, and this happens irrespective of whether the value is the same in the right and left collars, or the inverse (as is the case in Antonia).

The behaviour is not what I expected, and I wonder if I am missing some fundamental point here. 
  **
**
P.S.
I'm  using P6.