I find the first thumbnail in the first post more interesting, partly because it shows a relatively small rod parting the wind for a boxy building shape, but more so because it shows actual smoketrails to support the theory. Also, that the rod worked better when closer to the box.
This seems to work similarly to the way a canard wing affects and enhances the airflow hitting a main wing behind it on an airplane.
The void behind a car is large, and to fill it with a boattail, the tail has to be large enough to fill that void.
In contrast, if a small shape before a car can act to part the air before it gets to the car, then it's kind of a canard fairing/vortex generator. What's nice about this is that a smaller shape is going to be easier to experiment with, it doesn't need as much support structure as a boattail, and with a minimalist triangulated frame would present less surface to a sidewind, hence (hopefully) less of a negative yaw impact.
Fun to think about.
