I think Lehpron is talking about vortex lift as used on Concorde and a few other delta-wing aircraft (e.g. Saab Draken). I am sure there is no "magic number" -- the creation of vortex flow would depend on airspeed, wing sweep, camber, the leading edge radius, and a lot of other factors.
IIRC, another very factor is the compound curvature that induces the spanwise flow to wrap around the leading edge and twist into a vortex over the wing. I recall reading that the S-shaped curvature on Concorde's leading edge existed to produce a smooth transition from vortex lift to attached flow.
On a highly-swept wing, if the spanwise flow starts to spill over the leading edge of the wing, wouldn't the airflow separation (on top of the wing) occur near the leading edge? If it did so, I don't think it would reattach, and this would leave the wing in a stalled condition.
Yes, the airflow does stall, insofar as it becomes turbulent and separates from the upper surface of the wing. However, the rotation of the vortex over the wing re-energizes the boundary layer and creates additional non-linear lift. The only problem is that it creates tremendous drag. The lift-drag ratio on a conventional aircraft on approach with flaps and slats extended is probably around 12. On Concorde, with no high-lift devices and an ogive-delta wing, it was about 4.
Concorde actually required more thrust at 160 knots than at 250 knots, because vortex lift generation puts you on the back side of the drag curve (slower = more thrust). On a "normal" aircraft operating on the upward-sloping portion of its drag curve, slowing down will reduce drag and the aircraft will stabilize again at the point where drag and thrust are equal.
But Concorde at low speeds was in a disequilibrium situation,
in that the problem is self-reinforcing: if you slow down a little, drag rises, so you slow down a little more, drag rises a little more, and so on. If you don't watch your thrust margin, you can end up running out of power to maintain level flight. The wing never "stalls" since it's already technically stalled, but drag increases so fast that the engines can't compensate. This is what happened to AF4590 -- with two engines out and the gear locked down, Concorde needed to be going at least 220 knots to maintain altitude. They never reached this airspeed.
Keynes is dead and we are living in his long run.