Good find. Note that the graph you have shown simply displays the difference in CFD prediction vs wind tunnel results (ie the error) - not the actual changes in drag.
Versus the standard rim (wind tunnel results):
- fully blanked - Cd change of -0.013
- 85mm annular outer ring - Cd change of -0.009
So the Jaguar results are different to the Tesla results (where fully covered wheels increased drag), which in turn are different to the SAE 2011-01-0165 results (where ventilated front and fully covered rear wheels gave the best result).
As I wrote in the other thread:
To make it clear, based on the research I have cited in the book, I am not saying that fully covered wheels are bad. What I am saying is that fully-covered wheels aren't always best.
I am guessing it's highly dependent on the car to which the wheels are fitted.