The manner in which you change the angle is what changes the traction, not the angle itself. When you adjust the angle on bar cars, you usually ALSO change wheelbase, ride height, chassis preload, etc. On leaf spring cars the shim changes ride height.

Picture what happens in a rear suspension. And mind you, I'm talking about making just minor pinion angle changes, and not correcting u-joints out of operating range. The driveshaft rotates the pinion, forcing the ring gear to rotate down, and the pinion tries to climb the ring gear and rotate the housing up. Changing the pinion angle does not add any additional rotating torque to the pinion - it only changes the place on the ring gear that the pinion applies it's force. It doesn't matter if the pinion is at the 8, 9, or 10 o'clock position - X amount of torque is going to try to rotate the ring gear X amount, and try to twist the axle X amount. PERIOD. It's no different than using a torque wrench - if you apply 100 ft/lbs of torque, does it matter if your handle is at the 3, 6, 9, or 12 o'clock position? OF COURSE NOT, so I don't understand why people claim it matters where the pinion is trying to rotate the ring gear from. Same amount of force - same E.T. I can't explain it any simpler, but I'm sure some people still won't get it. Still waiting for someone to show with some engineering logic that I'm wrong, but nobody from the last 10 times this was debated has stepped up.

This ain't my first rodeo - any chassis and engineering classes I completed were IN ADDITION TO, not instead of racing passes.