Date: 27 Jul 2001
Well, you wanted the reason, here it is, from the engineer.
In my 924 owners manual Porsche talks about how the 924 has a negative scrub radius. This is ideal for suspension geometry; what it means is that when a different longitudinal force is applied on the two front wheels (such as braking on a split-mu, or failed X-split circuit braking), the wheel with the greater developed longitudinal force (the higher-mu wheel or the one that's still getting brake force in the failed circuit case) will turn towards the center of the car.
What this means is that the affect to the steering will naturally counter the effect of the inequal longitudinal (braking) forces. The inequal braking forces, by themselves, will cause the car to rotate (yaw) in the direction of the tire with the greater force. In other words, without any effect due to negative scrub radius, the car would veer towards the wheel that still has brake pressure. However, that one wheel, say, the right wheel for illustration, will actually turn to the left due to the negative scrub radius. Having a negative scrub radius simply means that the point on the ground around which the tire rotates is outside of the centerline of the tire, relative to the car. The tire forces (generated at the road-tire interface) are of course acting through the centerline of the tire.
Since the forces are not acting through the centerpoint of rotation, there will be a torque, or moment, around that point (we're keeping it in 2-D for simplification). With the force being applied inboard of the rotational center, the tire will rotate towards the center of the car. In the case stated above of this being the right wheel, the tire will rotate left around the steering axis, towards the center of the car.
Under normal straight-ahead driving, this would be balanced out by an equal and opposite force on the other tire, resulting in no impact on steering. When the braking (or tractive, for torque steer) forces are unequal, there will be a net force in one direction or the other, resulting in a steering change.
Going back to the failed circuit example, the tendency of the vehicle to veer towards the wheel which still has brake pressure will be countered by the tendency of that wheel to turn in the opposite direction.
Consider, however, if there is a positive scrub radius. The Lincoln Continental/Mercury Grand Marquis/Ford Crown Vic is such a car, with actually a very large positive scrub radius. In that case, on a split-mu surface under braking (this means different mu values, or friction coefficients, on each side of the car, such as if one side is on dry pavement and the other on ice), the car will naturally tend to veer onto the higher-mu surface (the dry pavement). The only way to counter that tendency is to modulate the brake pressure, which is why I know so much about this - I'm an ABS (and traction and stability control) engineer for Bosch, remember.
Anyway, back to the 924/944. The 924 has a negative scrub radius. Since the 944 (through early '85) has the same front suspension, I would expect it to have the same scrub radius, since it's determined by suspension geometry. However, you will all recall that the 944 has different offset and width rims. This may be the issue. The 924 Carrera GT, which has the same offset and width rims as the 944, also has a TT-split (front/rear).
That does explain it... to me anyway - kinda hard to pass on these concepts without a free-body diagram!