The larg horizontal turbines use aerodynamic lift that strongly depends on the angle of attack and the resulting vectors of motion. Any manipulation of the airstream changes this relationship, and it's alway down hill. Shrouding of any kind limits the potential energy, the bucket effect.
I guess you're right, I carried out an experiment with this wind tunnel I built:
https://physicsforums-bernhardtmediall.netdna-ssl.com/data/attachments/63/63801-87b7f3a5f83618e471314aa60073a5db.jpg , powered by a desk fan:
https://physicsforums-bernhardtmediall.netdna-ssl.com/data/attachments/63/63802-8eb3c548c8d1d3d9c24c88a46e71f050.jpg . The anemometer I got, is this model:
https://physicsforums-bernhardtmediall.netdna-ssl.com/data/attachments/63/63803-d9501a8fc0335f2ceb7eecd7a996a574.jpg with a propeller typed-blade. And is located in the blue cuboid casing between the 2 pyramids.
So it works by using the desk fan, located at the red pyramid, to suck in air through the blue pyramid, and the wind speed will be increased due to the Venturi effect, by 10 times, and pass through the anemometer, where the wind speed, or in this case, rate of propeller rotation, will be measured.
Thus to create a vortex/rotational/circular airflow, I made a "filter", by placing columns 45 degrees to the circular opening:
https://physicsforums-bernhardtmediall.netdna-ssl.com/data/attachments/63/63804-1f71128e0b16f14e4e18a1644adc9e0a.jpg , to make this:
https://physicsforums-bernhardtmediall.netdna-ssl.com/data/attachments/63/63805-8a66e32ad6252bbd7fe35d567839f5bb.jpg . When the air gets sucked in due to desk fan, in theory, the wind should be sucked in 45 degrees to the opening and thus creating rotational airflow(same clockwise motion as my anemometer's rotation to straight airflow).
I did 2 experiments, 1 without the "filter":
https://physicsforums-bernhardtmediall.netdna-ssl.com/data/attachments/63/63804-1f71128e0b16f14e4e18a1644adc9e0a.jpg , and the other, with. Here are my results
After measuring the wind speed(km/h) every 30 seconds, for 5 times.
The average wind speed of the experiment WITHOUT the filter, was: 14.48 km/h
The average wind speed of the 2nd experiment WITH the clockwise filter, was: 13.2 km/h
A 8.8% wind speed/performance drop...
I honestly have no idea why this is happening