I have yet to fly or hover a Harrier and I do not know its stalling speed when relying solely on its wings to stay airborne. In the hover it must be relying on thrust to stay up - there is no "Proper Lift" from the wings.
AF447 was no Harrier and lacked sufficient thrust to remain aloft without very considerably more thrust than TOGA. ( Three or four times ?)
I would guess that the aeroelasticity of the wing would have increased the effective dihedral, tending to help the lateral stability when descending.
Think of this as a "thought exercise", not to be attempted without further thought !