I think "balance spring" may need some rethinking. I have a little Z extension and my Rhino mounted at an angle, which exacerbates the issue, but I think its true for everyone with an extension or gooseneck where the top of the stick, or rather, its center of gravity, does not end up directly above the center of rotation. As its implemented now, I can apply a centering force for Y and -Y, but this doesnt yield a desirable effect. At "zero" position and even slightly -Y (forward) the stick will still want to fall towards me, and applying -Y force only makes it worse.
what I would really need ,is be able to tell Vpforce at what deflection angle the stick is naturally balanced (this could be very different from center position). No force should be applied there. As you deviate from this position, a centering force would be applied that is proportional to the cosine of the angle -and this force value would be the same on all axis, as it would depend on the weight/length of the stick. The applied force would be that value x cos (angle from balance point)