I’d like to clarify what Cd, the drag coefficient, is.
The drag of a thing, like a car, moving through a fluid, in this case air, is a complex function called…the drag function. It’s a part of the retarding function, but that includes friction and other stuff. In general, the drag function is of the form Sum n (v^n * Cd * Sa). (WordPress doesn’t have good equation fonts, sorry.
What’s going on is that each term is three components, the velocity bit, which is raised to a power, the drag coefficient bit, and the surface area of the thing (car).
The velocity bit is raised to a power, and that power goes up as velocity increases. This is the reason for the old 55 mph speed limit, because for a lot of vehicles, v^2 goes to v^3 or v^4 at 55 mph. However it doesn’t go up at the same velocity for all vehicles, and lower order terms don’t vanish. That’s why we sum all the terms together.
Regarding the Cd and Sa, imagine you’re looking at the car from the front. The surface area is like the area of the silhouette. It would be exactly the same if the vehicle didn’t have any curves or swoops or bits that make more surface area behind other bits of surface area, but cars do have such swoops, so the surface area is a bit complicated. Some of this complexity is approximated in the Cd. The Cd, drag coefficient, is how well something OF THAT SURFACE AREA goes through the air. So it’s a bit meaningless to talk about Cds by themselves because a Volkswagen Bus has a lower drag coefficient than many small cars. The VW Bus has much more surface area though, so it has higher drag. What’s more, the n part of that v^n above does seem to depend on Cd or Cd*Sa combined, so there’s complexity in there.
The drag function also changes for different air densities, pressures, mixtures, etc.
The point here is that just comparing Cd between cars is extremely misleading. The VW Passat has a lower drag coefficient than the old Supra but because the Supra had less surface area, it had a lower Cd*Sa. (That link uses frontal area and wrapped secondary surface area into the drag coefficient (I assume))
Anyway, it’s not just the coefficient you should talk about. It’s the combination, and honestly, the whole function is complicated. It’s an approximation of an unknown physical interaction which can be considered a hidden variable. What would be useful is to talk about drag at a speed and list a few data points, much like we list 0-60 mph and 0-100 mph when talking about acceleration.
Want to know why we don’t?
It’s extremely hard to measure and doesn’t make good copy.
Cd is easy to measure and makes good copy.