I think generic808 made a very relevant point that PERHAPS the OPer hadn't really thought about entirely. In the 2xCR123 size class, LEDs are really taking the lead, with well over 200 emitter lumens available and at least an hour or 2 of runtime. There's certainly reasons to appreciate an incan of any size or power, but sometimes, it's worth having your options all on the table.
Converting bulb lumens to SF lumens isn't as easy as you would think unfortunately.
SF lumens are not just de-rated for reflector and lens losses, but also take into account the "average" output that can be expected through the discharge. To convert LF lumens to SF lumens, it varies from lamp to lamp because SF de-rates some more than others. I believe this is a product placement/marketing strategy, so it gets even more complicated.. To convert LF to SF, divide in half, give or take 20% depending on the lamp. and because LF lamps are rated at a specific voltage, some perform in the real world better than their data sheets suggest, and others slightly worse... there are a lot more complications when you start throwing Li-Ion powered options into the mix, which drive some lamps much brighter than their labeled output.
For example:
an EO-9 compared to a P91.
take the 380 lumen rating of the EO-9 and divide by 2, 190 torch lumens. And in reality, this is probably a pretty accurate to life conversion. The EO-9 comes pretty close to the P91 when they are both driven on lithium primaries, and depending on the resistance in the flashlights being compared and the brand of cells used, the EO-9 will sometimes be reported as appearing brighter than the P91, the EO-9 also has a tighter hotspot, which can give it more useful throwing power for less power consumption, which often makes it appear brighter to people who desire throw. For the EO-6 rated 200 lumen, again, divide by 2, and you get about 100 SF torch lumens, and most people would agree that it is just ever so slightly dimmer than a P61 rated at 120 lumens but still a really great lamp. (darned close)... You have to keep in mind that it takes about a 30% difference in light output to really make a noticeable difference. Side by side, you can probably see differences as little as 10-15%, but in the real world, using one flashlight instead of another, larger differences are needed to "feel" any difference in performance.