I don’t think it makes enough difference to really count at the burner level. Several years ago there was quite a lot of spin inducing “vortex” burners (especially alcohol stoves) out there. Designs and actual stoves based on those designs were fairly common. Indeed, the BRS-3000 has a spin inducing burner at the flame/fuel/air mix junction. You can do similar with drilling a fine hole, then reaming it sideways with a somewhat larger polished needle on alcohol burners. This may effect the overall heat production/heat transference slightly, but not enough that anyone has been able to measure it while camping that I have heard about.
While some stoves (the discontinued Reactor for example) use flame to mix with their mesh/fill to provide a large amount of thermal energy (unusually high IR, but also exhaust gas convection,) the biggest advantage is overlooked. A dual layer pot with a turbulence inducing corner about the lower 20% of the pot. Still, the large CO emission as reported by Roger C. means there is some inneficiency in the heat production, possibly due to flame quenching, possibly due to insufficient oxygen at low flow velocities (when the stove is turned down ), possibly a combination of both or other things.
Anyway, however the heat is produced (WG, Kero, canister, alky, etc,) there is always the notion that the heat MUST be transferred through some substance to the target, water. I am ignoring IR here because this is radiation like sunlight or x-rays and follows more quantum rules. Laminar flow is great for clearing used materials from the system using traditional conductive/convective rules(exhaust gasses), basically, allowing heated gasses to flow up and fairly fast along some surface rapidly. Turbulence is induced where we want to hamper this movement (corners, vortex inducers, etc.) I don’t believe it matters one way or the other where the turbulence is induced, before reaching the pot, or, after reaching the pot. We are looking at side effects of inducing the turbulence. Less efficient combustion for example using the Reacter as an example. Using a full, tall, thin “Beer Can” pot means good laminar flow of heated gas along the sides for the full length of the filled pot; these achieve good efficiency of heat transfer also. Both low laminar flow rates and high laminar flow rates can be shown to be efficient for boiling water. A heat shield/wind screen can do both by redirecting the turbulence back to the pot. Well, all have been shown to be efficient mechanisms for boiling water.
By increasing turbulence you also increase disbursion of the heated gas: less efficient. But, any fins or disruptive elements added in the flow also pick up heat adding to the effect of heat exchanger. It works. I am sure you can come up with example where laminar flow decreases the efficiency of heating a pot.
Generally we always look at heating air as well as heating the pot. So, we have two materials in between the heat source and the water. Increasing IR with a flame can decrease efficiency because IR travels in a statistical 360 degree spherical pattern. Not real great for directing heat to a pot bottom without a heavy duty reflector (which gets hot and re-radiates IR…)
So, it seems we want a blend of these methods, both the inefficiencies of IR to gas in a flame and the direct IR against the pot bottom (allowing all else to escape) and the convective heat transfer of hot gasses to the pot which can be directed by design. In any wind, we ant more IR, in any calm we want more convenction. Thin highly conductive pots provide better conductance than thick less conductive pots. And so on. Again, different strokes…