With B12 the rear gets lower in comparison to the front and that means that there is more weight on the rear, making the front lighter and of course making the rear more planted but the front gets more understeer.
Lowering the rear ride height (relative to the front) won't cause the rear end to carry significantly
more weight. Don't be confused by the fact that forward acceleration causes rear end squat as well as rearward weight transfer, the two aren't inextricably linked, rather they are two somewhat seperate affects of the acceleration.
Statically raising or lowering one end of a car relative to the other will
move weight (represented by the CG of mass) forward or rearward, but only to a relatively small degree (even if the change in relative ride heights were to be quite substantial). It will be measurable in single digit millimetres, and of itself
I can't see that this small change in static longitudinal weight distribution will have a significant effect on the handling or steering etc.
Out of curiosity (and to quantify my assumptions...) I made a CorelDraw diagram. According to the diagram, if we assume a chassis with a wheelbase of 2595mm (156 sedan wheelbase), with a CG height at 600mm (about the same as the top of the tyres, generally a reasonable rough approximation of CG height), and a 60% forward weight bias determining the longitudinal CG location (typical FWD), then a 30mm lowering of rear ride height causes the CG to move backward by only 7mm (note that the CG also lowers by 12mm). This represents only a 2.7% change in the front / rear static weight bias (with a 30mm rear ride height lowering).
This is not to say that changing relative front / rear ride heights won't affect handling and steering, but I can't see that it would be as a direct result of altering the longitudinal weight distribution so slightly. The change in CG height
will be more significant, as will the associated change to the height of the rear vs front geometric roll centres.
Lowering the rear GRC (as will occur with a reduction in rear ride height, quite possibly significantly more than the actual decrease in ride height), will decrease geometric rear roll stiffness, so will reduce lateral weight transfer at the rear (while causing lateral WT to increase at the front even if front roll stiffness is unaltered). This will tend to increase understeer, especially initial understeer because geometric weight tranfers are 'instant' as opposed to 'mechanical' weight tranfers which are 'slow' ('geometric' being the % of weight transfer which occurs via the suspension geometry, and 'mechanical' being the % of WT which occurs via the springs and ARBs).
IMO it is probably counter productive for the rear ride height to be reduced if the front ride height isn't (likely to be understeer inducing, as well as degrading responsiveness to steering inputs and change of direction), so I do agree with you that you should try to avoid it happening. I just don't agree with the notion that it is a product of any change in static weight distribution...