If you need something for your late Christmas list, add an IR thermometer. My first one was $150, but they are $30-ish now. Look at ebay, J C Whitney, etc.
Here's the take-home message (I learned this from the first human to (1) quantum-entangle more than 2 photons and (2) find an exception to the Heisenberg Uncertainty Principle - my wife and I each have brothers smarter than ourselves): clean metal surfaces are MIRRORS to infrared radiation. That is bad, because heat your pot could absorb will be be reflected away.
With your spiffy, new, IR thermometer in hand: fill your pot with hot water. Check its "temperature" with the IR thermometer. If it is a "blackbody" (good), then it will read the water temp. If it is reflective of IR radiation (bad), it will read room temp. Maybe your treatment made it "black" in the infrared. But maybe not. Test it compared to actual black paint (BBQ grill paint or auto-parts-store paint for an engine block are both high-temp paints). Look at my tests from 13 months ago and I found that white, black and red paint were all "black" in the infrared.
The tabs seemed too big and too low. Big because you're trying to energize only the air film very close to the pot. Low because they were in the very hot gases. You want to "stir up" the somewhat cooler air. I'm guessing there was still a visible flame at that point (most HX was visible+IR radiant transfer) and therefore, the IR radiant HX continued further up the pot. Also, if those tabs fit within your windscreen, your windscreen could be a fair bit smaller. I'd guess you are diluting the flue gas temps with excess ambient air and that will definitely hurt HX.
One approach is "time to boil". That's good but subject to a lot of variables (room temp, stove temp, water temp, etc). Exhaust gas heat is another thing to monitor. If exhaust gases are small in flow rate and cool in temperature, you have captured more heat from them.
But a few careful experiments can often blow the best theories right out of the water.