My take on this...
As the servo-saver moves, the more it travels through its arc so the less lateral movement you get from the steering arms. As the arms get shorter, the less lock you get for a given travel of the servo saver.
Therefore, the issue here is not one of Ackerman, it is one of unequal steering from left to right. IF this is corrected by altering the end points, the overall effect is that you have less lock overall and therefore less steering overall.
Foam tyres do not have a single contact patch. Despite the scrub of the offset axle and inboard kingpin, foam tyres work because each part of the tyre is able to generate its own slip angle. With that in mind, the amount of lock you can achieve will dictate the amount of steering you can achieve.
One short link that limits the lock you can get is more of a problem than any effect on the Ackerman angles you set. Ackeman is actually set by the steering blocks, not the steering links. What is affected is the amount of lock you get from each wheel which looks like Ackerman, but it is actually just differences in among of steering lock.
Too much play is made of the Ackerman angles on cars with foam tyres. Since the tyre does not behave like a rubber tyre, it is the amount of lock applied to the inner and outer wheels that translates into different steering feel, not the Ackerman angle. In reality, a 12th car's cornering behaviour is more affected by the instantaneous steering centre than by the Ackerman or differences in lock from side to side.
The reason a steering servo must be in the middle is to maximise the love you can get from each wheel. The total amount you can achieve is dictated by the shortest link, so having each link equal increases total lock. Yes, I know you dial it back on the rate, but try driving a 12th car where the servo is offset 5mm one way. Once the lock circles are balanced on the Tx, the overall lock will be reduced a lot!
Congratulations to the guy who can tell the difference of a 0.005" off-centre position of his servo. Firstly you couldn't measure that accurately enough and secondly there is so much play in the whole front suspension that it could just as easily be a fraction more slop on one side of the whole front suspension, or one ball joint with a fraction more slop than the other. 0.005" is about the thickness of a human hair - forgive my scepticism...
