Brett from Japan has done some work on using cones on a canister stove, I am unable to find his postings on the topic in his 1400 post, but from memory the cone did not come down below the top of the canister therefore the canister does not over heat.
I am making a low profile remote canister stove for a cone style windshield and while trying to understand how a Cone works I did some testing with a canister stove under a cone with temperature probes through the sides at different levels, the cone only came down to the top of the canister, the first test had no holes at the top. I found that the hot gases stratified with the coolest layer on top, this stayed that way even until boiling was reached, I could not make it over turn.
The one thing that I did notice was there was a much higher temperature around the base of the pot than there is if there was with no cone.
I drilled some small holes around the top of the cone and repeated the tests, I found very little change in the above results (as you pointed out boundary layer thickness stopping the flow through the holes) so I made the holes bigger and more numerous re-tested and so on until I started seeing the stratification breakup, what I did notice is that the temperature at the base of the pot was higher in all tests and the temperature increased up the sides too as the bigger hole improves gas flow, all this is good for better heat transfer in to the water.
I did not try oval holes and thanks for the tip and also thanks for the figures of how thick the boundary layer is, I have measured them at the base but not the top.