Some famous race car designer (can't remember who) said something close to, "He who makes the best set of compromises creates the best design."
The practical point is: stay focused on the overall, and don't get obsessive about any particular feature or particular design criteria. I'm afraid I violated that principle. However, I was saved (from the grave) by Allan Staniforth. I'll explain, because we need to learn from each other's mistakes.
I was focusing on keeping my front roll center at 1" and minimizing toe in/out on bump and droop. Because we are mostly using donor parts, we don't have total control over suspension geometry. Certain values are fixed, like the location of the tie rod on the front spindles and the location of the upper and lower ball joints. However, one of the best tools we have to work with is the location of the suspension pivot points on the chassis, which we build ourselves.
So, I kept moving those to minimize the values mentioned above, but lost sight of their effect on other parameters like camber change of the front tires with body roll. When I was doing a simulation of a left turn with body roll, it became apparent camber change on the outside wheel/tire was only slightly less than the chassis roll angle, and it was also positive - not good.
Long story short, in re-reading some passages on camber change in Staniforth's book, "Competition Car Suspension: A Practical Handbook", I found the issue. I had let the upper and lower control arms become too near parallel, making the front swing arm length (front view) become too long, which has the effect of making camber angle change close to the change in body/chassis roll angle. So, at 3° of roll to the right, my chamber change was around 2.5° in the positive direction.
Luckily, I do a lot of simulation, so I caught the issue before I welded anything in place. Over time, I've developed a list of things to monitor, but front swing arm length was not one of them even though the software I use does display the value for you. It's now in my list. Interestingly, it is one of the first values set in Milliken's design procedure (p. 627), but they assume a clean sheet of paper design, and all values for the front spindles are set by the designer, which is not our situation, typically.
Attachment:
Suspension Targets.jpg
So, it is truly "back to the drawing board" and finding a better set of compromises. I thing front roll center height will be one of them. Trying too hard to keep it at 1" off the pavement led me to some (unintended) poor compromises elsewhere.
Cheers,