Oh no of course not. If I did do it I’d be mixing a couple packages worth of JB and casting additional heatsink in-situ.
Theoretically, the heatsinking and a driver swap is all that would be needed to cross the 10,000 lumen threshold. Devide 10,000 by 18 emitters gives just over 555 lumens from each emitter. Let’s assume I can get an uclp lens, and because of the small reflectors, I’d only have a 10% loss. So I need about 620 lumens from each emitter. This means each emitter needs about 2000mA, if I keep the normal non-DTP mcpcb’s. Wired in a 5-4-5-4 configuration, I would be drawing ten amps from a stack of 4 cells. But, there is a fifth cell. That means an extra volt for each block of emitters, something they could tolerate quite finely. And I’m willing to bet that the extra cell’s worth of voltage would maintain that ten amp draw long enough to heat-soak the head, similar to the manner that if you DD a XM-L2 off a 4s-NiMH stack that you get more than 6 amps.
Plus, the side switch means I can use a momentary driver. (hopefully the light already uses a momentary there). So there’s no losses or risk of melting a clicky with the absurd wattage this thing would flow.
I still want to do the Trustfire I linked in my OP because I have a Trustfire boner, its green, not absurdly huge, and will have a minor throw advantage over this monstrosity.