I've done some "oil drops" airflow studies for this opening. The flow separates pretty much at the edge of the windshield/roof and it appears that quite a bit of turbulence is present on the roof in front of the duct inlet. The top edge of the scoop is in enough clear air to pickup airflow and create a small localized small high-pressure area. This is just enough to force some airflow into the duct.
From the windshield edge, the drops migrate reward for about 2 inches and then start to flow in all directions for about another 3-4 inches. The drops tend to stop moving for the next couple of inches and then at the entrance of the duct they begin to flow towards it from a couple of inches out.
I did a row of drops at the other end of the duct in the engine room and there is airflow out of the duct so it is moving air.
Below is a picture of a Porsche in a wind tunnel. The roofline is similar to an SLC in that the window angle is fairly steep and then transitions into a much more flat roof. You can see the separation at the windshield very clearly. Picture our SLC roof scoop right in this spot and you can see what I am talking about.
One of my future experiments is to form a forward-facing extension and attempt to pick up higher pressure airflow at the edge of the windshield. I am pretty sure this will work but it will be ugly.
The second picture is a good example of the extended scoop duct idea.