Here's the latest from my warped little mind. It's only slightly related to the thread, but I don't want to start another thread for this bit of silliness.
What if the quarter elliptic leaf springs were solidly attached to BOTH the frame and the axle housing and the back end of a single upper link was also attached solidly to the housing. In other words, the only true pivot would be at the front of the upper link. Obviously, you wouln't need a Panhard. But, what particularly intrigues me is that ride height could be easily changed by adjusting the length of the upper arm. As a design consideration, the upper arm would have to be long enough to minimize pinion angle changes and its pivot point so located as to yield the desired amount of antisquat. It would be mounted slightly to the right of the driveshaft. While cornering, the springs could just be allowed to twist (as they presently do, but without the help of the rubber bushing at the eye).
(Yes, the spring would have to be custom. When compared to the more common single free end, the spring rate would be doubled, and, for a given area inertia and deflection, the stress would be doubled.)