A longstanding reality of LEDs is that they're more efficient at lower currents. Within otherwise identical formfactors, running multiple LEDs at low/medium currents nets both greater lumens and runtime than a single LED at medium/high current.
The obsession with throw confounds me. As much dust, humidity, pollution, and light pollution is in the air in my region, going out to >200m away means steadily increasing backscatter to the point that you're looking at the beam path itself as much as you are light reflected from the target. I find myself using floody lights pretty regularly (increasingly these are triples and quads); the throwers are borderline useless for day-to-day tasks and used rarely. Living in the suburbs, I feel like my needs are fairly representative of a big slice of the market - dog-walking, task work, ceiling bounce, illuminating a small work space or work area, occasionally using turbo to briefly illuminate an area some tens of meters distant.
I think the big/deep reflector concept has endured more out of inertia
(incandescents needed them in order to extract meaningful Cd from what were historically low-lumen sources for most of their existence) and ill-informed perception of utility on the part of the market - high Cd impresses but is not of so much use in the field. Much like how high-power LEDs have been displaced from the general lighting marked in favor of greater counts of low-/mid-power LEDs, I can see parallels in the flashlight market - less heat per component is a desirable design goal from an operating and reliability perspective.
I mean with these new bright LEDs, going forward with new pocket light designs, I think simplicity and efficiency will push the design of triples and quads from ever gaining a foothold in mass produced pocket lights. We won't see them, and if they do appear, we won't want them by then.
Pocket lights? Probably. 100-200 lumens seems to be adequate for the keychain or something single cell outside of li-ion formats. Little benefit to be had there when even the lowest-cost most humble power LEDs going into those designs can handle far more power than those designs consume. Short of some unusual bumps in LED efficiency and breakthroughs in cell technology I don't see those dynamics changing much.
But the CPF demographic doesn't buy these mass-market designs. Nor does the mass market buy lights fueled exclusively by li-ion cells priced mid double-digits and up.
And outside of pocket lights I don't see the tri-/quad-/pent-/hex-/sept-/oct-/etc-emitter trend going away. Supporting commodities are readily available from multiple producers in semi-standard formats - optics, reflectors, MCPCBs, drivers. Heck,
Home Depot sells multi-emitter flashlights at prices the masses find acceptable.