Try a 16 x 8 Wide or a 16 x 10 narrow, the plane can handle the speed, and supressed RPM lowers engine temperature... (I did not test that with the Boxer, it has had the same prop from the teststand until now) but all other engines remain cooler if I prop heavier).

Also: how much runtime does your engine have in total, and after conversion? From new, 20 hrs is nothing. The rings are seated, but friction still is not at its lowest, Then I have also seen conversions that although broken in on glow fuel, still needed 10 hrs or so to stabilize. Especially when Castor oil was used previously. So chances are it will improve over time.

On the subject of "temperature", some considerations:
Personally, I hold 145 as the absolute limit, I refuse to go hotter than that, and throttle back preferrably at 135, simply because I don't want to risk an engine. I do not have any hard indication (engine damage or other signs) what the REAL limit is, and I most surely want to keep it that way since new engines are harder and harder to get, but I have seen sudden RPM loss around that temperature in the first two fourstroke conversions (.52 and .91, both ASP, both broken in on gas, back then still running 5:1 fuel/oil) and I never went beyond that.
For clarity: Engine Temperature is an indicator for, not a cause of engine damage: Modern oils can do their job easily up to about 180 degrees, Pistons easily survive 250 degrees and exhaust valves even handle 600 degrees or hotter.
It is that last one that is the stinker: Gasoline autoignites at around 400 deg C, under compression even sooner, so that valve can serve as source of ignition. Cooling of the valve happens via metal-to-metal contact between valve and seat, and THAT is why CHT is important. As long as CHT is within limits, not only will the rest of the engine most definitely be OK, but it means that the exhaust valve won't cause pre-ignition. And pre-ignition is the killer, causing thermal runaway of all parts forming the combustion chamber.

I have tried all kinds of "cooling mods", some worked, most didn't It is most definitely possible but the difference between working solutions and non-functional ones really are very small.
Saito's in general handle things much better on account of their construction (cylinder and head in one single casting, dissipating heat away from the head way better).
Although I do not have any experience, I would expect the Enya AAC construction to also perform better due to the liner being Aluminium. The steel liners of ringed conventional engines forms a heat barrier. It has little use trying to cool the engine better, focus on cooling the head.
A "cooling plate" between rockercover and head helps to a certain extent, but no miracles should be expected and results vary with the aerodynamic shape of that plate. Internal control of the airflow under the cowl also has marginal effects, but in combination, they can give unexpected results that can really be surprising.
Just as an example: The .91 was always a bit of a "heat challenged" engine, and there was not much difference between running the cowl or not.
When I cowled the engine, I attempted to direct the air entering the cowl as efficient as possible towards the cooling fins. This basically did nothing. That air guiding was mounted to the motor mount, and had a gap of maybe 2 or 3 mm between itself and the cowl for ease of removal of the cowl. No difference noticable in cooling effect.


I fitted a cooling plate between head and rockercover, that did "something" (it was measurable but nowhere near significant.

I closed the gap between cowl and air baffle, this did "something" but similar to the cooling plate, barely measurable.

Then I folded the cooling plate a tiny little bit outward and DANG.... THAT made a huge difference. I have had it at WOT for 1,5 minute and not hit the temperature alarm,

What I am trying to say, is that optimizing cooling most definitely is possible, even with our "undercooled, designed for glowfuel" engines, but the devil is in the details. Large mods and high time-consuming constructions can appear to be virtually without effect, tiny changes suh as that fold can make all the pieces of the puzzle fall into place.
What I think happened was that I focused on getting air to the engine, instead of getting air AWAY from the engine. In hindsight, the same thing occurred when I baffled the radial in the FockeWulf.
The baffling, and the cooling plates did "something" but the gamechanger was the spoiler in front of the air-exit, creating a low pressure area promoting air extraction.