Mad,
My description of SAS was far too simplistic but all I was trying to convey was that the SAS is not an autopilot. I fully appreciate that there are variable gains within a modern autopilot (I have been trying to find the optimum numbers for MRA4 for a while now!). SAS is not trying to control the aircraft to a pre-determined set of conditions whether they are flight path management, speed management etc. There is no great distinction between inner loop and outer loop functionality. When comparing to the autopilot, SAS is puely inner loop as the pilot's have no means of monitoring the performance. The autopilot outer loop functions can be monitored by the pilot by looking at the flight director or performance instruments, there are no such 'visual' outer loop SAS performance terms.
The SAS has been designed to provide flying qualities of an aircraft with an apparent CG that is forward of the actual CG. It also has aileron position compensation terms (prevents aircraft pitch-up with large aileron deflection), compensation terms for thrust lever movement and flap, gear and undercarrigae selection.
I am not trying to make out that this is new technology (I am unsure of equivalent systems on civil aircraft), what I am trying to convey is the time frame from conception to flight has been excellent.
The split of engineering between Woodford and Warton does not prevent any difficulties with the flight test and engineering aspects of the project. What has been very clear is the few times we have operated the aircraft out of Woodofrd, aircraft generation is far superior at Woodford than Warton.