Okay, so here's the intial plan:
1) get a thermocouple on the head
2) run the engine on the ground, with cowl, using 10%, read the temp
3) Switch to 5%, repeat an compare temperaures.

4) Modify the cowl with better ducting
5) run the engine on the ground, with improved cowl, using 10%, read the temp
6) Switch to 5%, repeat an compare temperaures.

So, whats good temperature and whats a bad temperature?

I'll probably switch to a uniflow on pricipals but for these tests that should not affect the results.