That said, SM did fix the symmetry stuff I complained about, so a thumbs up to them for that.

Symmetry issues should not be so difficult to tackle and when of type 1a, 1b, 2a above easy to fix as long as the drift is of a lesser order of magnitude as the mesh size. Move vertices to the average mirrored position and done. This could be done even in Poser as a script tool. Issues of type 2b above could be signalled and added to a vertex group 'unpaired', and left for the use to fix.  A first step would be to make the search tolerance in Poser a quantity the user can define. Not sure if it is in not already in the settings files somewhere.