We put a new canopy seal in the IV-P. I fiddled with the cabin pressure valve to try to reduce or locate  a squeal. (My Grainger smoke matches don't light, with reduced partial pressure of Oxygen at altitude) I noticed that pulling the cabin pressure control partially out, reduced the cabin inflow, as expected, but also dropped the manifold pressure by about 4 inches with a loss of about 50 hp (estimated). This was at flight level 230. I had not previously noticed this large effect on manifold pressure, but it could demystify performance changes in previous data. Who knew the cabin pressure valve affected the "throttle" setting :) It makes sense because the previous cabin pressure just dumps overboard at the firewall. Wondering why this is set up this way.... Why not just close the valve to the cabin, when not needed or wanted, and leave the manifold pressure in the manifold??

  In general, I noticed others have similar issues with the cost of repair of the Dukes valve. I would rather redesign the system, using the on-board instrumentation and computers that already know the static pressure and the cabin pressure. I hot climates I prefer to minimize the heat entering the cabin, and maybe just use 4-5 psi relief valves to distribute air flow. I think my system where all the inter-cooled air routes thru the firewall to under the back seat was designed by Eskimos, who always need max heating.