So we're basically talking about two different things here:
1) limiting how much droop the suspension has in order to keep a preload force on the springs at all time, also can be used for keeping parts from over-extending and doing bad things, such as ball joints going past their intended travel.
2) more extreme limiting of droop travel (sometimes so extreme that there is no droop travel at all) to enhance the vehicle's handling.
All of my posts have been in reference to the latter. "Droop limiting [on the front] gives less understeer." I have wondered many times of this would be legal for setup limited street class autocross cars. I've never heard of anybody doing it.
Here are some quotes from Z:
Getting back to the issue of "droop-limiting gives less understeer". Here are two factors missing from the above explanations.
Factor 1.
First the basics. Assume linear rate springs. Assume only just droop-limited, so no preload. Assume typical FF wishbones, so long and almost horizontal virtual swing-arm. Don't think about "roll-centre" (a dodgy concept for independent suspensions). Instead think about line from wheelprint to swing-arm IC - call it the "n-line" (normal to wheelprint up-down movement), or "control-arm-force-line" (as per Mark Ortiz). For the moment forget about ARB's, dampers, and rear suspension.
The droop-limited pair of springs will have twice the roll-stiffness of a normal setup (ie. for given roll moment a normal setup's 2 springs compress/extend "X" and roll angle is 2X/Track radians, while on the droop-limited setup only one spring compresses for roll angle = X/T.) So for given cornering G (and forgetting rear susp.) the droop-limited is stiffer and only rolls half as much. Stiffer should mean more understeer, but half roll angle means less adverse camber change so less understeer. So far this is as noted in above posts.
BUT!!! When the normal setup rolls there is no change in chassis centreline ride height. When the droop-limited setup rolls the centreline ride height DROPS by X/2. This drop in front ride height lowers the slope of the n-lines (= lower RC). Lateral load transfer (LLT) is a combination of spring forces (= elastic LLT), control arm forces (= virtual swing arm, or n-line, or kinematic LLT), and damper forces (= viscous LLT). When the n-lines slope down from the chassis to the wheelprints there is kinematic LLT off the inside wheel and onto the outside wheel. BUT! When n-lines slope up from car centre to the wheelprints (= below ground RC) then kinematic LLT is ONTO the inside wheel and OFF the outside wheel.
So for typical FF wishbones a lower front ride height means LESS kinematic LLT and hence less understeer (from this factor). Lower ride height will also, typically, mean more negative camber on outside wheel, again less understeer. Of course, the car is only interested in the "bottom line". It takes account of all (hundreds+..) factors, their sizes and signs (+ or -), then adds them all to decide whether it is going to under/oversteer. So really you also have to know what the dampers, rear susp., etc. are doing...
Droop-limiting is like fitting falling rate springs (the force/deflection curve bends down to the right). Rising rate springs (curve bends up to right) are great for absorbing bumps, and are often used as a justification for rockers (use linear coils and make rising rate with rocker geometry). Falling rate springs pull the car down during roll, but rising rate lifts the car! A lifted car (higher CG) is a BAD thing, hence most setups that work well are either linear or perhaps falling rate (at least in roll, "third spring" acts only in bounce so can have rising rate). (This is another justification for rockers out the window.)
Factor 2.
A lower front ride during cornering will pull the nose closer to the ground. Formula Fords aren't supposed to be "aero" but a lower nose will probably create some suction under it. Since this is cantilevered out the front of the car it may (?) increase front download AND reduce rear download, hence less understeer.
http://www.fsae.com/forums/showthread.php?4047-zero-droop-behaviour