I just finished metering my Manker E01 and will be posting the results in the next day or two. It was a test in Med mode, as my goal is to test all my multi level 1xAAA lights in medium to compare to my Fenix E01. I used Energizer Ultimate Lithium cells in all the tests as that's what I use in my EDC lights.
I assume you are talking about this graph:
So, the sole purpose of this series of tests was to take all my 1xAAA lights with a similar output to my Fenix E01 and see if technology has advanced far enough to surpass the runtime vs brightness advantage that's always been in favor of the E01. As I've stated dozens of times over my time here on CPF, the three biggest advantages of the Fenix E01 over other lights of that period (for me) were 1) it's ability to fire on a nearly depleted cell, 2) the constant output runtime and 3) it's ability to take any abuse you can through at it.
I find the results interesting, but it should be noted, that the E01 used was an original, when they 1st came out. I thought I've read that Fenix has since upped the brightness and lessened the runtime or something along those lines.
I found that data very useful, thank you for generating it and sharing. I took the time to characterize my Manker E01 in the only way I can, which is to measure the voltage and current into the driver with a calibrated power supply. I don't have a light meter, so I use the power supply to tell me if/how the light attempts to regulate. The data I gathered for medium mode for simulated battery voltage of 1.7V down to 1.01V was that the power into the driver is very constant. From this I conclude the light output is also constant. Two example data points are for 1.7V input, the current is .0.56A, so the total power is 0.095W; for 1.01V input, the current is 0.098A, so the total power input is 0.099W, or only 4% different than the 1.7V input. Somewhere below 1.01V the light likely comes out of regulation and the output falls of quickly as your graphs show. The driver puts out a tiny bit of light with 0.85V input, and is out when the input drops to 0.8V.
What was more interesting was the behavior on Low and High mode. In low mode, the current into the driver roughly follows the voltage into the driver. That is, for 1.7V in, the current is .007A, so the total power input is 0.012W. For 1.01V input, the current is 0.003A, so the total power input is 0.003W. No regulation going on here, the light dims as the battery voltage decreases.
High mode is even stranger. From 1.7V to 1.5V input, the light appears well regulated. But between 1.5 and 1.4V. the input current jumps by 30%. As the voltage continues to go down the current goes up for a bit, then at 1.2V start decreasing. It makes me think this light is well suited for a lithium iron battery on high, but not so much an alkaline or even Nimh rechargable.