What is known, so far, is that the aeroplane crashed as it was returning from acceptance test flight and that's it. Everything else is uncertain. We don't know if it stalled, whether it pitched up and what was the active control law before the crash.
There is a stall warning and a stall recovery procedure on the A320
Stall warning is available in alternate and direct law, when there's no alpha-prot to stop the aeroplane from stalling. However it is possible to get false stall warning in normal law, if AoA probe is damaged. I'm not sure if this FCOM warning can be interpreted as "single (out of three) damaged AoA probe can cause false stall warning", but if it can, it may imply that voting rules are not the same for AoA and speed/attitude. QF72 final report will be very, very interesting reading indeed.
Procedure commonly refered as stall recovery (full power, decrease pitch, wings level) is actually recovery from approach to stall. The significance is that it is initiated as soon as stall warning goes off and therefore before actual stall takes place. If one allows the airliner's wing(s) to stall fully, this procedure does not guarantee recovery, or at least recovery that doesn't include overspeed/overstress/ground contact and its success is largely dependent on all-up weight, centre of gravity and luck.
In the event the voting process goes haywire, is there a way to manually get to the point where the pilot(s) have the full control as you describe when you have all three units disagreeing?
It is possible to get into direct law with some simple switching but the procedure is strictly unofficial and, as I'm unaware of any occurence of voting process gone haywire, i think it's unnecessary. I stand to be corrected, of course, if some incident/accident report shows otherwise.
System design very well explained except for the case where two similar but erroneous inputs are sensed by the ADRs (from Pitot/Static system).
We have three indepentent systems, each required to be very reliable. Suddenly two of them simultaneuosly fail in the exactly same manner and their false outputs are diverging from real values with the very similar amplitude over time. I didn't take this scenario into consideration as my opinion is that chances of its occurence are virtually nil.
and you could get your Alternate Law 1 too...(and have real control of your aircraft, without loosing a great deal of protections
I beg to differ. One loses every protection in Alt law
(EDIT: except G protection, thanks for reminding me). What Airbus euphemistically calls "reduced protections" are really not much of a protection at all.