Streamlined flow and turbulent flow are fundamentally different.
All flows must be parallel to a solid boundary at molecular distances, but streamline flows are parallel at slightly larger distances and if some sort of flow visualisation is used (say smoke in air or dye in water etc) time lapse photos show that all particles are following the same paths which are parallel to solid surfaces even if they are curved.
For turbulent flows, there are time dependent fluctuations existing at all times, and while average flows may be predictable, instantaneous flows at any location may be different in both direction and magnitude from those observed at an earlier time.
It is possible for flows in a boundary layer to be turbulent while flows outside that boundary layer can be laminar. Bernoulli's theorem which allows estimation of pressure fields for laminar flows ( P + 1/2 rho V**2 = constant along a streamline) shows that Pressure P is reduced if fluid velocity V is larger. So over the top of an aerofoil where V is larger, P is reduced, and the net effect is that the higher pressure P below the aerofoil compared to that above provides lift.
In a turbulent boundary layer (aft of the separation point on an aerofoil upper surface in a stall), flows are unpredictable AND Bernoulli's theorem doesn't apply. The reduced pressure is much less pronounced; and lift is much reduced.
SB