Might the variance in specs be due to static camber changes causing SAI to be different from the Kingpin angle, which would be the same as SAI only at 0 camber?
As to scrub radius affecting feel: I have found larger scrub radius having increased kickback and negative impact on feel. Increased caster and the resultant trail on the other hand gives a very nice feel, to where the steering gets lighter as grip is diminishing.
If you look at the scrub radius and it's orientation to lateral forces, the amount of feedback due to lateral forces will be minimal and varies a lot with steering angle, and feedback from longitudinal loads(ie braking and bumps) will be much greater, Trail is the longitudinal distance from the contact center to the steering axis, and is oriented much better to have an effect on "feel" for cornering forces.
I believe this is the reason almost all front wheel drive vehicles and most modern rear wheel drive vehicles have near zero or negative scrub radius. If the scrub radius is negative, the feedback from longitudinal loads like braking, bumps, and in the case of FWD, drive forces are self correcting. when the brake force is higher on the left side with a positive scrub radius, it will pull the steering to the left, which is the same way the forces are already pulling the car. With a negative scrub radius, the higher braking force on the left side will pull the steering to the right, helping keep the car going straight. The same principal works for bumps and drive forces. If you've ever driven a FWD car that has much wider wheels than stock, offset in the negative direction(to the outside) increasing scrub radius in the positive direction on a FWD car, you can notice torque steer is almost unmanageable.
The only problem I see with negative scrub radius is that with a positive kingpin inclination, there is negative jacking forces. This is why it would be ideal to have very high(positive) offset(rim to the inside, hub center to the outside) so you can actually get a negative scrub radius and a negative kingpin inclination. It takes very wide inner rim shells and very narrow outer rim shells, and a narrow spindle setup with as little as possible offset from the ball joints to the hub surface to achieve this though.
One way I wanted to try and achieve this would be to use a small solid axle spindle, kind of like a sprint car spindle mounted to a short stub axle that is welded to a larger spindle/upright that doesn't turn with the wheels, kind of like the spindles on an early Chevy or Ford independent front suspension. You could put the negatively inclined kingpin spindles inside the brake rotor hat, right next to the wheel mounting surface.
kind of like this: