I wouldn't doubt that different cars react differently to lowering. All that I'm saying is that any practical lowering (>3" above ground) reduces aerodynamic drag (if all other parameters are left the same).
By shear forces I mean frictional forces and frictional forces do play a role.
E.g. frictional forces increase with a 'rough' underbody (which also increases the size of the boundary layer, which then also increases pressure drag).
In the meanwhile I have received the following book: "Hucho - Aerodynamik des Automobils" by T. Schütz et al:
On page 201 it shows pressure and frictional drag for different Audi models (recent models) and frictional drag causes about 10% of the total drag (I believe all these models have a smooth underbody).
There's relatively little information regarding lowering in this book:
On page 169 it has a general statement that Cd is reduced (not just frontal area) by 0.004/10 mm of lowering (doesn't say what this is based on and what range it is valid for).
On page 316 it shows data from another book (Janssen und Hucho 1973) of 4 different cars.
The Porsche 914 (with the lowest Cd of all cars) shows the largest Cd reduction due to lowering before a not-named-car and a VW van. Only one car shows a Cd-increase after an initial Cd-reduction (Citroën ID 19).
On page 317 it shows the only recent lowering data (Audi Q7).
The Cd*frontal-area is reduced by 9% at 50 mm lowering and is increased by 4% at a 24 mm raise.
The data is almost on a linear curve and stops at 50 mm lowering (so much more aerodynamic drag reduction could probably be achieved with more lowering).
However, I don't quite trust the old data on page 316 since this must have been measured before they introduced moving floors in wind tunnels and a moving ground leads to a significant reduction of frictional forces underneath the car.
(As I mentioned before, the frictional drag between two moving surfaces is
12 times higher than of just one moving surface. If you read in chapter 2 'Analytical Navier-Stokes Solutions' of this book:
http://homes.nano.aau.dk/lg/Lab-on-C...20lectures.pdf you'll come to this conclusion if you solve equation 2.14 for differential pressure due to frictional forces and do the same with equation 2.58. )
And this factor 12 is based on laminar flow without any boundary layers, whereas in a turbulent flow regime you have two boundary layers if the ground is not moving as opposed to just one boundary layer with a moving ground (factor 12 and one more boundary layer).
The Citroën ID 19 in the example above probably just started to touch the boundary layer on the wind tunnel floor at some point, which then started to disproportionally increase the frictional forces underneath the car which subsequently increased Cd. (In 'Race Car Aerodynamics' by Joseph Katz it is mentioned that the boundary layer on a wind tunnel floor has a thickness of approx. 100 mm - this boundary layer is missing with a moving ground or on the road for that matter. The fact that efficient ground effect race cars (without the use of fans) were not 'discovered' before the mid 1970's was because a moving ground in a windtunnel is an essential part for air to remain attached to the underside surface of the bodywork in order for it to generate a significant downforce.)
Back to the Hucho-book:
All data which is pitch angle related suggest that a lowered nose always reduces Cd and a raised nose always increases Cd.
For instance, the VW 411 had a Cd reduction from 0.45 to 0.42 with a -2 degree negative pitch angle increase and the Audi Q7 a Cd*frontal-area reduction of 5% with only -1.5 degree negative pitch angle.
So, if you choose to lower only one part of the car, you should lower the front or raise the rear but not raise the front.