Right. Good! So....let's take it even further.
In an arc'd turn we can say that we have minimized that rotational torque and maximized the turning effect on the actual skier. If that is the case, then in an arc'd turn, the tip and tail would need to neutralize each other. If there was negative steering angle on the tail, then they would not neutralize each other at all, they would collude together to create a pinwheel effect. They must both be positive and they need to be similar in value also in order to eliminate that rotational torque.
In a side slip you have also eliminated the rotational torque by ensuring that both the tip and the tail cause approximately the same amount of summed up steering angle to occur on each side of the boot(front and back). Steering angle is very large in that case, in fact its so large, with reduced edging, that its not steering the skier at all, its only producing a lot of skid. But the point is, that there is large steering angle in the front, large steering angle in the back, and the delta between them is going to be something very close to zero. Its not the sum, its the delta, that indicates the rotational torque.
Now then take the arcing case, We have reduced the rotational torque in this case also, but we have reduced the overall steering angle of the ski to as close to zero as we're going to get while making a turn. Whether its zero, or very very close to zero, it mattereth not in the discussion about whether there is ever any negative steering angle. The front half of the ski and the back half of the ski need to have a very small delta between them in terms of their local steering angles. If the front half is positive, the back half needs to be positive also. Otherwise there will be more rotational torque.