[Schulisco]

The sensor plate height defines the way the lever can make and related to this it defines the amount of fuel being injected. Saying that - the lower the sensor plate is adjusted, the more way it will travel based on the same amount of sucked air. In other words: the mixture gets richer on lower rpms and over the whole rpm band. On the other hand: the higher position of the sensor plate , the more lean the mixture will be, because the lever has less way to travel.
This relates directly to this is the basic adjustment setup of the CIS : If the basic setup of the CIS (my personal wording is "injection begin") is not correct, this also affects the mixture in the same way, but in finer resolution: The CO screw is basically nothing else than a fine tuning mechanism for the sensor plate height. The goal is now to find the right point that the injection at the cylinders begins precisely exact then, when the lever is barely lifted by the sucked air of a) the starting engine or b) to keep the mixture especially on idle as lean as possible and as rich as required.
The rest is pretty easy because there's nothing more to adjust, because every component of the particular CIS system is adjusted to the engine. The curve which defines the fuel amount to be injected is "hard coded" in the air flow sensor housing and its components (fuel distributor, sensor plate housing, venturi size etc) to match the engine's individual requirements.
This explains why you can't simply replace a fuel distributor from on a 911 SC with one from e.g. a Mercedes. The curve of injected fuel according to the sucked air is mechanically hard wired in the mechanics of sensor plate housing, fuel distributor, injectors etc. This is a math function which in todays systems programmed in a ECU.

You can imagine the way the sensor plate lever travels (and causes a particular amount of fuel being injected) as a "sliding window". This window is simply the direct relation of lever travel and fuel amount. This is basically not linear, because the mixture requires to be richer on higher revs. If the window is either too low or too high, the mixture always is partly correct only, but never over the full rpm band! This explains why some engines then will start good, but having poor performance and really bad fuel economy. Others start really bad, but run better, have bad low rpm behaviour etc.
On top of that we have the WUR (and AAV) to enable cold start and running which enriches the mixture by lowering the control pressure. And on US SCs from 1980 model year on we have the lambda control, which also influences this "mechanically programmed curve" to make the catalyst working.

Thomas