Te effects of torque steer I was seeing on my own car were most likely due to positive caster and large scrub radius.
As we all know, positive caster adds stability in a straight line, and even more so during heavy braking. But positive caster also has the exact opposite effect when tractive effort is applied at the front wheels. The contact patch being behind the steering axis. If there is also a significant scrub radius (either way), this can lead to trouble if the front wheel wights change from side to side.
In low powered FWD cars it is not that serious, but with a lot of power, the tractive forces are much greater.
What I was getting was the car started to gently rock gently from side to side under acceleration, creating a slow cyclic body roll. The weight being transferred from side to side at the front, causing the car to weave.
The front was also much raised in height during this acceleration.
This weaving and rocking would be self reinforcing, gaining in amplitude. that, and the very light steering, becoming quite disconcerting.
Trying to correct for it only made it far worse.
Best to just hang on grimly and back off a bit....
I believe the cause was the combination of caster and excessive scrub radius, the lateral weight transfer and tractive effort combining to self steer the front wheels. This steering causing more body roll, which created more self steering.
The drive shafts on this particular car are of identical length, but I have never understood the often quoted theory that unequal driveshaft lengths can somehow steer the car.
I am working on the design of a four wheel drive hot rod right now, and plan to run zero front scrub, and only about three degrees of caster.
All quite interesting.