I saw some attempts, lots of time and achieved Nil. My 2 cents (and only a current opinion) is that getting air out is the key.
Good to see all the pictures
I have spent huge amounts of time and effort on this and can summarize
1. Get temp probes sorted - under plug rings changed depending on each plug removal, also had cold junction in the engine area, finally had head temps 90-120
changes to Jabiru between head push in probes, temps jumped to 120-140, ended up with CAE probes and temps stable at 130-140 - same cowls and ducts (different engine)
2, As per Jabiru USA tech tip on top deflectors, this really works but takes time and doesnt help left to right bias
Learning from my trials and pics here that cowl exit is a BIG factor, notice pics here it is much larger and Jabiru have several versions of lower cowl and seems the big ones work.
3. Oil cooler inlets stuff up flow.and mean process has to be looked at again.
3, EGT, spread is a massive problem with every airframe being different. Actual temps are sucessfuly adjusted with jets or leaning valve.
Air box and inlet NACA do induce swirl and RAM pressure, which Bing seems especially sensitive to.
Newest hole inlet is next to be tried or at least running with cowl hose off. A small amount of carb heat really makes a difference to EGT spread, not because of heat but disruption to flow. A pod filter would do similar.
Almost a black art but critical to get engine working right
A catch up at YNRM is possible, I have a place in hanger there