It may have something to do with the rigid ring which usually keeps the opening of the windsock open. I have a small non-rigid windsock made of ripstop nylon which has no stiffening ring around the upwind opening and this rarely stays properly inflated - it often shows a tendency to collapse in on itself.
It may also be that the initial assumptions about the airflow speed increasing are incorrect or over estimate this effect. I agree in a venturi with a constrained airflow both up and downstream of the narrower section there will be an increased flow rate and reduced pressure in the venturi. However the windsock is different - the air both up and downstream of the windsock is not constrained. At the upwind opening of the sock there may be local slowing of the airflow due to turbulence around the mouth of the sock, and some air may spill around the outside of the sock in eddies. There could then be a net increase of pressure inside the sock which is sufficient to keep it inflated. I bet if you placed the nozzle from the end of an air-start hose inside the windsock and blasted a jet of air through it to re-energise the flow inside the sock, it would suck it closed tighter than the sphincter of the Air Transat pilot who lost both engines of his A330 in mid-atlantic.
Has anyone out there ever measured the pressure inside and outside of a windsock? If so that is very sad and you need to find yourself a good psychiatrist as you need help.. .-----------------------------------------------. .Pobody's nerfect