[Trevor Booth]

Chris L
I respectfully suggest you re-check your calcs, If the upper trailing arm is not affecting toe in what is it doing to the geometry.

Clayton,
The forward end Point D1 is behind the upright casting and a long way from B2. Check out the 4 element sketch below. Point B2 is swinging about the same radius as D1 but it should be a different radius to D1.

Russ,
Geometric interference "winds" up the system and then can suddenly release.
ie. in the 4 element. B and D are rotating about line A-C but at same radii. B is tending to rotate about D (in the plane of the wishbone) this winds up the system and usually stresses the trailink link A-B until it deflects. The trailing link will resist deflection up to a point (and still be straight) all of a sudden it will deflect. It is this sudden and instantaneous deflection that causes a like change in the geometry and un predictable handling.

In the event you have a straight clevis with a reversed lower A arm, slacken all the bolts in the upright and watch what happens. In the event you have offset clevis, point B will adopt the true position on A-D

Lowering the rear roll centre may not necessarily promote more understeer, it depends on its relationship with the front roll centre.

I cant conclusively say that lowering the front inner point on a parrallel link sytem would induce toe in, would need to do the maths but I suspect it may well be cancelled out,,just a thought. You really cant look at any element in isolation, it is a system all working together (hopefully :-)))