In the article they say that the coefficient of aerodynamic drag was from CAD.
I would speculate that they meant to say CFD.
According to Mercedes-Benz's Dr. Toby Woll,Direct Numerical Simulation in their supercomputer will render a Cd which is within 1% accuracy of a full-scale wind tunnel model.
If DNS was performed for the Probe X,then their Cd 0.15 ought to be accurate.
Hucho showed us half a dozen ways to get Cd 0.15 from 1981.It shouldn't be much of an issue.(if we keep mirrors off the things!)
Here's GM's 1983 Aero 2002,Cd 0.14.Similar,with truncated Kamm tail.
Volkswagen's 1981 truncated-tail 'Flow' body,Cd 0.15
