Hi Takata,
At this point there appear to be two paths to take an Airbus with properly operating computers in Normal Law outside normal flight limits. One path involves bad airspeed data leading the computers astray, but also implys that the RTL position would be different than that found on AF447's VS.
The second path can start with properly functioning airspeed and still lose control. It relates to factors that take control of the aircraft away from the computers.
Do you do any test flying? Take a look at the Gripen accident I am posting a link to. You only need to watch the first run-through. Do you understand the cause? (And I have posted this link previously)
YouTube - Saab JAS39 Gripen Crash 1993
You can read a translated summary of the accident here:
Summary of report on JAS 39 Gripen crash on Aug 8:th 1993 in Stockholm
It appears that the original full accident report is not online.
I am not promising you that this is THE cause for the loss of AF447. The data gained from the FDRs will (hopefully) tell us what really happened, and it could be stranger still. This type of pathway to loss of control would tend to cause the aircraft to depart in an oscillating (or dynamic) manner. (More to post on the subject later)