BreezyRacer wrote:What actually occurs is that the true roll center moves, because the springing variance between the loaded side and unloaded side forces a roll center location change dynamically. To a similar extent you get a similar effect with bump/rebound settings, though then only with initial weight transfer.
First, I have to thank you. Your posts have made me think about compound front springs, long bumps rubbers, & the like. They are used widely in modern road vehicles &, much less widely, in competition vehicles.
I think of the road vehicle long bump rubbers as a "poor man's bar", because efficient arb's are difficult to install in a front engined vehicle (& they can be a "noise" transmission path), and because they ensure that the vehicle will understeer at the limit. They do have an issue, because tracking several road vehicles through their lives has demonstrated that the bump rubbers used in road vehicles take a permanent set, & therefore become less effective, with time.
For what it is worth, I concluded that a balanced steady state turn requires a particular CPL distribution, regardless of the mechanisms used to achieve that distribution. However, those mechanisms do determine the deflections of individual suspension elements. Thus, for example, suspension geometry will determine the proportion of the roll moment reacted by the springs and arb, the remainder being reacted by suspension links. An increase in "roll centre height" implies that less of the roll moment will be reacted by the springs & arb, suspension deflection will be less, so the transition to steady state will require less time (typically shortening the steering time constant). The disadvantage is that the sprung mass will tend to "climb over" the loaded side, because the centre of gravity will move away from the ground plane (or, at least, move less towards the ground plane).
Compound springs set so that the "helper" springs are collapsed at static ride height will, in a steady state turn, cause the loaded spring to deflect less than the unloaded spring. This will also move the sprung mass centre of gravity away from the ground plane, acting very like a high roll centre geometry.
To simplify the task considerably, a vehicle that is well damped (& therefore has good control over contact patch load variations) will have dampers that are, on average, matched to the springs. Increase the springs, & the damper strengths must be increased to maintain optimal CPL control. The problem with compound springs is to decide how to optimize damper settings. Almost inevitably, the vehicle will be under-damped in compression & over-damped in extension. The conundrum is not minor. I owned a vehicle once that had to be slowed down over an uneven section of road because long front bump rubbers caused loss of control. Also, I have seen clear air under the tyres of a (non aero) competition vehicle excited by relatively modest vertical inputs on a rig, again because compound springs were fitted. Usually, for competition vehicles, reducing the spring stiffness "split" (or replacing the compound springs with linear equivalents) has improved on-track performance.
I have to conclude that compound springs are a mixed blessing. They can help one aspect of performance, but they compromise others (increasing sprung mass c.g. height in a turn is not a good idea by itself). Overall, I believe that sensible geometry, linear springs, appropriate damping "styles", and an efficient arb usually yields the best solution, particularly for high performance vehicles. How that relates to the present topic is anybody's guess.
Undoubtedly it's pretty complicated, but I still appreciate any dumbed-down explanations you can give me! Thanks! 
Hope someone caan shed some light on it soon.
