I have been meaning to graph the contribution torsion bar vs jounce bumpers to draw conclusions about ride stiffness. I was too cheap to buy a a line pot for a data acquisition unit, so I never got anywhere.
It would be very easy to do for someone with access to basic mechanical instruments at a college, tech school, or employer.
The question to answer: What is the spring rate of the jounce bumper?
With that info you can make decisions about modifying it to get a lower overall effective spring rate. Maybe drilling a hole down the center of the bumper solves all problems and no one needs new torsion bars.
Pic attached demonstrates the concept of overall spring rate. At some distance of suspension travel, the jounce bumpers make contact with the control arms and the overall spring rate of the suspension is increased. This is the blue shaded portion of the graph.
If anyone knows spring rates of the torsion bars or equivalent conversions as if it was a coil spring, that would help as well.
Also, I have always understood the spring rate of torsion bars to be linear, therefore cranking the keys does not change ride quality, all it does is lift the suspension. This is a common misconception from people with cranked keys because the upper control arms kit the service stops on down travel, making for a crap ride. The concept of "preload" for linear coil springs and torsion bars resulting in a rougher ride has no merit. This is not true for a progressive rate coil spring.
Fingers crossed someone here is a physics teacher and measures and generates a graph accounting for the torsion bar, and 2 jounce bumpers corrected for motion ration.