<<As for big cowlings and little spinners making no difference, I
would point to the large number of Unlimited air racers that have achieved
large drag reductions through the use of highly oversized spinners vs. their
stock configurations.
Eric Ahlstrom>>

All good points, Eric.  I didn't mean to imply that oversize inlets made
"no" difference, just that in the grand scheme of things it might not be the
worst of sins.  Certainly the "Rare Bear" is a good example of how to
correct for the oversize inlet in the original design - however, the speed
increase certainly didn't correlate to an equivalent reduction in flat plate
area.

A few things to consider, none of which are exactly new:  We usually use the
Coefficient of Drag, Cd, as a way to normalize data from different shapes
and it represents the impact, or stagnation pressure distributed across the
frontal area of the shape.  A flat plate actually has a Cd of about 1.5 and
the reason it is greater than 1 is that the air flow off the edges is
outward, essentially blocking airflow and making the plate look bigger than
it is to the air stream.  Something to remember when pointing an exhaust
pipe 90 degrees to the airflow - the plume itself causes drag.  As I recall,
a bullet has a Cd of about 0.1, meaning that 90% of the drag (even more
compared to a flat plate) is eliminated by a smooth front.  But if you
reverse the bullet so the flat end is forward and then apply an optimum
shape behind the flat area you can reduce the drag at least as much.  To do
this the overall area has to increase as you need to provide a smooth radius
outside the blunt face, making it look a lot like a radial-engined airplane.
The two have equivalent drag, but the one with the flat face has a much
larger "stagnation" area.  One point is that stagnation is a required
phenomena at the front of any shape and it theoretically exists at only one
spot.  Accelerating the air from that point to the maximum velocity is
relatively easy to do efficiently as the pressure is dropping as the
velocity increases, tending to maintain laminar flow and minimizing losses.
The hard thing is to decelerate that air back to zero velocity without
incurring losses.  In that case the pressure is gradually increasing, trying
to push the air in the boundary layer upstream and eventually "tripping" the
air flow, which then separates from the surface.

Apologizing if I sounded overly simplistic,

Gary Casey