I've run a WA1111 with 2xIMR26500 successfully using an FM G4 D26 Sunlight and FM Deep Turbo C head in an FM 2x26500 SureFire C threaded host. Tail switch is a McClicky.
Expected tail current is in the 3.8A-3.9A range. Even if you have only 100mohms of parasitic resistance in your flashlight system (switch, spring contacts, socket contacts, etc), that translates to a parasitic voltage drop of about 0.4V. So you can be somewhat close to the approx instaflash point of 8.3V from Lux Luthor's destructive incan table for the 1111. But clearly, the 1111 can definitely work without flashing. As FM says above, it depends on the host and how much parasitic resistance you have in your light. One thing to note about Lux's tests. He directly powered his lamps from a bench power supply using thick gauge cables to a Kiu socket. His setup had only about 20mohms resistance. So you have to factor in your system's resistance as well. The flip side is that power was delivered to the lamps slowly (effectively, a very slow soft start) since Lux had to turn his power supply knob to increase the input voltage. Thus, for a flashlight application, a soft start will also help.