If you study enough of this, one look under the cowling and the shortcomings are obvious.  I've got Hoerner's Fluid Dynamic Lift and also Fluid Dynamic Drag.  Although it is still the best book for finding out about every aspect of drag (or lift for the companion book) it is pretty dry.  It isn't as easy to read as some of the newer books like Daniel Raymer's book "Aircraft Design: A Conceptual approach" and Darrol Stinton's "Design of the Aeroplane".  I'll bet somebody has some really extensive research on this waiting to be discovered so I doubt I'mbreaking any new ground here.  Even a 3-4" radius would make a big difference I suspect.  The flow exiting underneath would be largely separated, however the lower pressure created by the flow outside the cowling would be lower and smoother helping remove the air more effectively.  This should reduce the boundary layer thickness, I suspect.

 

One of the articles I read (I can't remember where now), stated that for years, many designers were adding more inlet area to improve cooling but getting very little extra benefit.  Then someone started reducing the size of the inlet but increasing the size of the exit.  This reduced frontal drag while increasing the amount of air flow.   It was from that article that I started thinking more about improving the "quality" of the cowling exit. 

 

One of the funnier aspects of shaping the through hull exits was that some were so effective at creating a low pressure region that at high speed, I actually has some developing lower pressure than ambient thus actually sucking air through the fittings!  I first witnessed this on my first solo race to Hawaii while trying to sleep below.  When the boat started surfing (under autopilot) and it started making this wierd kind of moaning sound then if it really broke loose on a fast surf, it sounded like a Kirby vacuum cleaner!  I had to stick a plug in the galley sink to make it stop!

 

Dan