delmar, thanks for your precise observations. Since you're using basically a jim woods fancy feast stove, which in my tests is both the fastest and hottest burning and least efficient, that's an excellent counter example to my Ion type sgt rock tests, which focus purely on efficiency.
You found also what I found by accident, lifting the entire screen up provides the exact amount of air flow required, which is what I aimed to match with holes, by simply calculating the area of air gap lifting it creates, then putting that into holes, ie, making the area of holes match the area that the lifted screen created. I found this during my fairly exhaustive testing process, which I did in many stages, height of screen, lifting it to create air flow, determining optimal height off ground (I believe 3/16 to 1/4, depending of course on the overall screen diameter, I assume the bigger the diameter, the less gap is required). I used paperclips too, but found them too annoying to actually use for the final design because they slip, fall off, etc, but you can create the same result quite simply by first determining the optimal bottom air gap height, then simply leaving 6 or 8 nubs on the base of the screen and cutting out the space between them. 1/2 to 3/4" per nub should be strong enough, and it's a lot easier than cutting out those air holes, which is a real pain.
Here's my how to make alcohol stove windscreen instructions.
Note that I did not update that with the second screen information because I didn't test it adequately, but maybe I'll do any update and do some tests with second flashing external screen and a top mesh screen for wind.
I like your idea of making the second screen a permanent part of the design, that allows you to always have the main screen lifted, and, more important, to have the air gap as low as possible on the main screen surface.
Just as an aside, I believe a lot of the stove issues are all about air turbulence, and something I realized when trying to match the air gap area with holes was that a bottom gap creates a completely smooth area of incoming air, like water pouring down a big pipe, all around the outside edge, evenly, as opposed to discrete jets of air shooting in. My thinking re lifting the screen vs holes was that on dirt the screen might sink in a bit thus making the gap smaller whereas holes would always be above the ground. But sometimes I overthink things, making the gap on the bottom is a lot neater and easier, for sure.
Another thing worth noting is that if you have holes on only half the circumference, which is a common design, that's what I did first too, the air flow is not even, and is even less efficient, it's quite striking testing this stuff how dramatic the differences are, and it's also striking that lifting the screen so it's got a gap all around its circumference is so obviously the most efficient that the overall design should be based on that idea in my opinion, that's why most of the efficient setups out there use really big air holes, it's pretty simple actually.
To avoid speculation and incorrect theories, here's my thinking on the second screen, an image would show it better, but I'll try to put it into words. Since I'm assuming some behaviors of air flows and I may be wrong, the best way to actually determine if this idea is wrong or right is by testing both methods, inside/outside second screen in wind. If there's one thing I learned during testing, it's that air / alcohol /screen works in weird and very specific ways that can only be determined by empirical testing.
Let's assume we make the main screen with the bottom air gap, I put my holes as close to the bottom as possible, but simply having it raised is better because that places the gap as low as possible to the ground, but I didn't figure out a clean way to do that so I went with holes. My thinking is: with the second screen on the outside, what happens is that all moving air approaching the screen bottom is stopped, and the air the screen chamber pulls in basically gets sucked down from the top of the second outer screen. This makes the inner chamber work like a cylinder, with the proper air flow, but with protection against air turbulence, since the air is getting pulled in. Obviously, things to test with this approach would be the overall diameter of the second screen, the outer one in my case, to see is there is also an optimal spacing between the main screen and the outer screen, and an optimal height for the outer screen. That type of testing is much harder though.
The problem I would expect to see by placing the second screen inside the main screen is that the moving air is going to create air turbulence in the burn chamber, which should in theory have a strongly negative effect on stove efficiency. That's in theory, however, if you are using a fancy feast type stove, those are hot, very very hot, the hottest I tested, fastest boils by far.
I was also taken in by the 'intuitive' idea that a screen should be taller to trap the heat against the pot body as far as possible, however, unfortunately that simple could not be verified empirically so I had to admit that sgt rock was right, that's not actually how the burning / efficiency works with this setup, which is when I started to realize that the stove/screen/pot/gap between stove/screen were one system, not a few, which is of course when I hit the efficiency numbers I'd been looking for for so long.
Determining optimal gap for the fancy feast stove would require the same testing, but it's actually very easy to do since all you have to do is take a longer piece of flashing, raise it 3/16 or so, then clip it together to form varying diameters, until you find the most efficient. 3/8" to 1/2" for that sounds about right to me.
I didn't do much testing with the external second ring because my first try, a strip taken from a heavy foil baking pan, worked perfectly, and removed the issue caused by fan moving air across the bottom, so I didn't test it further, but now I'm wondering if I should make a real second screen, out of flashing, and do a bit of experimenting with it. Problem is, of course, that experimenting takes forever, heh.
If it's true that screen over the top helps with the air turbulence however, that is significant, would be interesting to see.
By the way, if you used some type of framed pack, and aren't alone, two packs pushed against each other to form a V shape should create a decent wind break for a stove, unless it's very windy.