What's wrong with the geometry of those VGs? And how can you tell they're wrong from a photo when the teams have spent hours in CFD / wind tunnel playing with them before mounting them on the car?
If you are more fortunate than others, build a larger table not a taller fence.
They're intended to interact with the rear tire, the fact they don't race them means they're not having the intended results, and this serves as more or less a correlation study for further development.
Your limited CFD and windtunnel time is better off going into more important/long lead time parts than it is something you can just roughly guess what you want it to do, 3D print some in a day and just stick 'em on the car to see what it does in practice.
