To your point about the front of our roofs resembling the leading edge of a wing, I'd have to agree with a caveat. VGs aren't on aircraft wings to help out at cruising AoAs. They're there to stave off stall and maintain lift and control power at extreme angles of attack. A car never really sees anything other than a static angle, so having them in the flattest area doesn't make sense. You could put them ON the windshield in an effort to help the turning onto the roof. That would be an interesting study. Putting them on the roof near the front, though, is going to be less effective than further downstream where the closeout angle is more severe.
On to the link you posted. There's a lot of things going on there that make me question the quality of the CFD work. The comparison images are almost worthless since the writers didn't use consistent color scales, and detail is lost since they used banded coloring. I could go on, but it's not entirely relevant. Just be wary of blindly trusting that study.
That said, their results aren't necessarily wrong. They very well could be reducing lift the amount that they quote. That doesn't have anything to do with what VGs are intended to do and would be more of an unintended result. Their sole job is to re-energize the boundary layer by tripping it and creating vortices, thereby keeping flow attached. In the case of the Evo, their purpose is to reduce drag and get more air on the wing.
If you speed up the flow over the rear of top side of an S30 the pressure is going to drop and lift will go up. That's just physics. It's why 911s had to go to whale tails to keep the rear ends down.