As simple as I dare in the presence of a proper chassis expert, my take is this:-
Take a look at the stock cars toe rod, its inclined down wards, the outer pivot is below the inner.
As suspension compresses the arc causes a realtive lengthening of the toe arm, increasing toe out.
The reverse would be true for the inner wheel as it unloads.
In effect both rears will steer slightly in the opposite direction to the fronts, aiding turn in.
Saab do something similar to aid understeer but they do it with differering rubbers and therefor compliance in the bushes I believe.

Take the lowered chassis and no other mods:-
The toe arm is now no longer inclined out and down.
Firstly you have to shorten it to regain correct rear toe values (adjustable toe rods).
At about 40mm rear drop you find the toe arm about horizontal.
So NOW when you compress the suspension, the arc doesn't effectively lengthen the toe arm, it can't any more.
The toe arm travels upwards above horizontal now and so can only effectively shorten again.
This I presume would indice a small degree of toe in, not good and acting opposite to design.
The shim stack on the end of the Megan rods alters the couple once more and also returns the rod to about the original 2-3 degree downward incline.
Probably over simplified but that's what appears to happen

Without altering the pivot points / lengths I doubt you can do much about all this really.

 

This is how it seems to work in my mind aswell.

 

If that is correct, then I agree - compression causes toe out. I don't have a standard car to go look at otherwise I would.

Okay, but that's an effect of roll, not an effect of compression due to acceleration/deceleration. We know the effect of the RWS is to steer more on trailing throttle and to give more stability (i.e. toe in) under power, and you can feel this at moderate speeds meaning it's something more than the usual effect of weight transfer. How does that fit in?

Which is the perfect position as far as bump steer is concerned (minimal horizontal movement for vertical movement). So why do people talk of increased bump steer with lowering?

 

1. It is correct.
2. Same thing. Throttle off, rear rises, wheels toe in more.
3. Not so simple, you would think that but study this new 'zero' position.....
The range would now be to shift towards toe in under compression and also toe in under extension of the suspension.
In effect with the toe rod horizontal it's a long as it can be in its arc of travel.
therefor it can only effectively shorten, always making toe in wherever the suspension is and not desirable.

HTH?

 

I see this is getting a revisit following a recent post, so to add..

Bumpsteer kits on the rear of a lowered car:
After a 40mm drop, the rear toe arms are almost horizontal again.
So further spring compression from maximum 'extension' moves the arm upwards - shortening it.
This will transition into a degree of nasty toe in, where the design intent was the opposite.
Returning the arm to a three degree angle or so by a bumpsteer kit resolves this.