*side note: the "low" pressure side of my cowling seems to be significantly charged with air trying to escape. You can see the oil door arching in flight, and the piano hinge cowl to fuse margin *seems* to expand a smidge from what I assume is air trying to escape.
All normal. What matters is (1) the pressure difference between the upper plenum and the lower plenum, and (2) how much heat is transferred to the air.
This chart maps out required deltaP at various altitudes and OAT's. The heat transfer is a built-in empirical assumption. Do better than the assumption and you'll need less deltaP.
Regarding pressure, a builder may (1) increase upper plenum pressure, and/or (2) decrease lower plenum pressure, either of which results in more deltaP. Increasing upper pressure is more desirable than decreasing lower pressure, as more lower cowl pressure results in higher exit velocity. Less velocity loss equals less cooling drag. For example, assume a deltaP of 8" H20 would result in the desired mass flow for a low CHT. In fast cruise you might have 18 upper and 10 lower, which will definitely make your cowl bulge. Or (given very poor conversion of available dynamic pressure, Q) you might have 10 upper and 2 lower. Cooling will be the same (both 8" deltaP), but the second case will be higher drag. The bulging cowl indicates a
good thing.
It's a double whammy. A cowl with poor conversion of Q won't generate enough upper plenum pressure at climb speeds, even if the lower cowl pressure was at freestream static. Can't get a deltaP of 8 when you start with 5.
Obviously another way to run hot is to have a bunch of leakage between the upper and lower volumes, as (1) it tends to equalize pressures, and (2) it lowers heat transfer (it won't meet the empirical assumptions in the cooling chart).
I've rebaffled, sealed up what I can, timing is spot on, cleaned up the exit area as best I can - all the usual suggestions here. I have an EZ-Cool flap ready to install, but it seems to me that it may only be part of the solution (useful only for climb out.) I've got some pressure numbers around somewhere, but they seem to suggest a less than ideal pressure ratio top-to-bottom (I need to find them.)
Glad to look at those numbers. And tell us the whole package; engine, compression, ignition type, test altitude, OAT, etc. Got pictures?