<<  That would be nice...  but if the plenum fits the cowl, nicely, why
not use
the cowl to seal the whole thing, instead of the extra weight of the
"plenum".>>

Extra weight! What extra weight?  Silicone baffle seal material is not
exactly light weight stuff.  My calibrated arm tells me they're about
even.  Seriously, if I were to put all the removed  parts on a scale, I
would not be surprised to conclude the same thing Brent did - the plenum
is lighter.  

I actually did consider using the cowl as the plenum top.  There were
two issues.  The first was how to transition the top of the diffusers to
the cowl without creating a step that would cause flow  separation given
sealing and engine movement considerations.  
The second issue was controlling leakage in a conventional baffle
system.  NASA did a study pertaining to the cooling of horizontally
opposed aircraft engines that produced some very enlightening data.
They took an Aztec engine installation and found 55% of the air entering
the inlets was being lost to leakage.  38% was attributed to the
flexible baffle seals and the rest was traced down to various sheet
metal to engine interfaces.  A hardtop plenum and RTV got things sealed
up.  Accepting a method of sealing that is so inherently lossy is not an
issue if you have an oversized cooling system.  I was going in the
opposite direction by reducing inlet and exit areas.    System
efficiency was now of greater importance and so I did not want to go
down the path of accepting a sealing method that has been shown to have
a very high leak potential.  Do leaks really have an impact?  If you are
already losing half your cooling air, you could probably drill a few
more holes in the baffling and not see a difference in cooling.  On my
first flight with the plenum, I actually had a leak.  One of the SCAT
ducts was not pushed up onto the diffuser quite far enough.  There was
no visible gap, but the internal spiral wire was not entirely up on the
diffuser so there was a short section of unsupported silicone creating
the seal.  You could just barely squeeze a finger between the diffuser
and the SCAT duct in that area.  On that first flight, the cylinder
closest to that leak ran 20 degrees hotter.     Do leaks really make a
difference?  In this case it did.  
While a conventional baffle systems will work ( I used one for 6
years), it is, IMHO, much easier to achieve a well sealed and efficient
system with an enclosed upper plenum.  The benefits of this efficiency
become more apparent, the faster we want to fly.

Chris Zavatson