I've run a WA1111 in an FiveMega G4 D26 Sunlight using 2xIMR26500 in a FiveMega host (with McClicky tail switch) with no instaflash. I've also successfully run a WA1331 in a Sunlight using 3xIMR16340 in an FM 2x18500 host body and Z41 tailcap.
Electronguru's data that you presented is confusing to me. I can't tell if he has actually tested the listed configurations or if he is claiming
simply based on nominal Li-ion voltage vs instaflash voltage as
measured and tabulated by Lux Luthor.
IMO, it depends. Lux's measurements are based on manually adjusting the input voltage from a bench power supply, through heavy gauge wires, to a bulb inserted into a Kiu socket. On the one hand, Lux's setup has probably the minimum parasitic resistance, and hence parasitic voltage drop. On the other hand, it is also the ultimiate "soft start".
So, how do those numbers translate to a real flashlight? Tough to say. A real flashlight could have enough parasitic resistances to drop the Vbatt to a safer level wrt instaflash. But even with some voltage drop, is the bulb ramp-up sufficiently gentle? Who knows? That can also depend on the statistical variation in the bulb filaments, as well as how close Vbulb is to Vflash.
It is clear, however, that the farther away you are from Lux's Vflash point, the safer you will be wrt instaflash (this shouldn't be a big revelation). The question is how far away is enough, since you probably want to balance safety from instaflash with high lumens output. Maybe secondarily, also consider two more variables -- filament life and heat generation. It seems to me that folks are using at least a 1V cushion below Vflash as a rule of thumb. For example, see
Lux's practical guide to using his bulb tables.
So what low resistance mods, if any, are in your 2D Mag? Sometimes having some higher level of parasitic resistance can actually help you.
