Yeah, it's been done. I made about twenty of the little 1 liter ones with various indents in the bottom, finally settling on a radial 5 ridge format. This was about 15%-17% more effective at garnering heat, but looses effectiveness with slower heats. At about 20-25 minutes to boil 2 cups of water it seems the pot radiates heat as fast as it absorbs it. Best time/fuel ratio was about 10-12 minutes with that pot even though it was only about 10-12% more efficient than a flat pot. The actual cooking surface was increased about 20% on the bottom only. (About 1/4" deep ridges makes it look like a "Fresnel" style lens, 'cept made of aluminum.)
Just aneal the bottom then carefully tap around a jig to get it to corogate. Then burnish with a harder metal burnisher (cleaned often to prevent metal/metal sticking.) I continue to use one.
No, it isn't all that difficult to clean. The 5 ridges were sized to a finger so I *can* scrub them out if I cook scrammbed eggs or something elaborate, like a muffin/bread on the trail. A couple teaspoons of olive oil does well at keeping stuff from sticking badly, unless you burn stuff (too much heat.)
A nice experiment if you want to try it. I drew a bunch of extended conclusions with it based on about 50 tests. It allows heat to radiate out as fast as in. Good for cooling a pot, too. It performed identically to a slightly larger, but flat pot, leading to the conclusion that heat exhanges only appear be larger diameter pots to a heat source. All were done with a Cone and a 12/10 burner. All pots started life as a standard K-Mart grease pot. Interesting, a single "dimple" in the bottom was only slightly better than a flat one at about 2-3% - within measument tollerance - for a 3/4" dished bottom.) I assumed it was from heat "puddling" in the bottom cavity rather than flowing out smoothly. It just adds insulation, rather than transfering heat to the pot.
2,3,4,5,6,7,8,9 ridges were tried with anything more than 5 becoming impossible to clean if you cooked anything. The difference between 1-2, 2-3, 3-4 was greater than the difference between 6-7, 7-8, 8-9. I settled on 5 as the optimum for cooking(evenings) and boiling water(mornings.) I destroyed a few pots trying to make the ridges deeper (1/2") and settled on 1/4" as the optimum depth. One was 3/8" that worked well but it was difficult to clean, though not impossible. It gets quite thin, though I eventually cut through it while slicing some cheese.
The aluminum can be "stretched" so it does not add weight. Mine weighs about 3.25oz including the top...not bad for a 1 liter pot. Origonally designed for an alcohol stove, it works as well with Esbit, WG or canisters. At about 15-17% better efficiency, it works on all stoves except for very slow heats. I usually set my heat to get around 9-12 minutes for 2 cups and it burns about 4-5g of WG in a SVEA 123r or about the same with canister gas and a Coleman F1 (overheats in the cone) or Caffin's stove. I never chased it down any closer. Works is works...
Go for it...it definitly works. Takes about 30 minutes to make one.
The radial design was just easy to work with. Look up metal forming over moulds. This explains the metalurgy. Other designs are certainly possible.