Thanks for the responses guys.  OK, so let's redesign the system.

1.  One thought is to simply disconnect the scat tube from the air filter box
in front of the number one cylinder.  Block the filter opening so air no
longer passes through it.  Let the unattached scat tube suck in warm air
underneath the cylinders.  Provide a suitable screen so that nothing but air
is sucked into the tube.  I've had cooling trouble in the past with taking off
using filtered air because the cooling air goes down the filter inlet instead
of over the cylinders on that side.  It's a noticeable CHT difference vs.
taking off with ram air.  With this alternative, air would be sucked from
below the cylinders further improving cooling flow (right?).  I also need to
do the air flow calcs to see which way air will tend to flow in the scat tube
at cruise, since I can't close it off (I can only close off ram air).

2.  An alternative might be to install a spring loaded warm air door
somewhere.  But if I'm flying with the ram open and freeze the throttle body
air sensor tubes somehow (do we know of any cases of this actually happening?)
I will have to move the control to close the ram air off anyway.  In which
case would not simply drawing unfiltered are from the aforementioned scat tube
be just as effective (and less work and small parts risk)?

3.  I could move a larger filter box down behind the ram air opening and
install a warm air door downstream of the filter.  The goal would be to give
up as little ram air pressure boost as possible.

My air sensor (in the throttle body) is directly (line of sight) behind the
ram air opening by about 12 inches.  The current Y pipe is directly in front
of that, leaving about a 3 inch gap to the cowl opening (which is bridged with
scat tube).   That setup is not easy to change.  What operational restrictions
would you impose on using ram air in such a setup?  Is the risk just icing, in
which case, like carb guys, we'd worry about moisture and a certain temp
range?  Bugs?  There's four tubes there and blocking four tubes on a single
flight is remote.

If the point is, we need warm air, then I can do that very easily with number
1 above.  Is anything else called for?

Thanks for the help!

Ed de Chazal