I just realized that the LOD file extracts correctly the aero loads ONLY when the lifting surface is HORIZONTAL.
Whenever the surface is inclined or vertical it delivers the right X-Y- Z-forces as well as the right X-moment. However, the Y- and Z-moments are wrong.
Please, find in the attached file a simple test case that allows checking it
Basically it is a turbine-like assembly where the lift of each element add up to generate axis torque. The arrows in the next image shows the vector Cx*i+Cy*j+Cz*k (force coefficient vector) which is OK :
However, while plotting the moments coefficients with respect to the origin (Cmx, Cmy, Cmz), we get weird results.
Indeed the top view showng the moments of the horizontal wings is OK: the moment components are OK
But the moments on the vertical surfaces are not OK --> the Cmy and Cmz values are wrong:

For information, I ran several test cases with multiple geometries (from simple to complex aircraft configurations) with the same results. It is then NOT case-related but rather an internal bug. I'm using 3.46
Best
Miguel