If by "finicky" you mean something else, please forgive my assumptions. And if you would please define what you actually mean by finicky, I would very much appreciate it. My FW3A lights all lock out with the slightest turn of the head or tail, but I turn the head. And the lights always work once retightened.
If the tailcap or head rotate at all - including the application of
any notable torque to the clip - the switch will either cease functioning or
latch closed.
The primary design failure for the FW3A is that it effectively depends on a
line fit for both the inner and outer bodies WRT making contact to the PCBs. At the price point the FW3A is made to this is not feasible
(the parts can neither be made to near-zero tolerance nor can the inner body be post-production fitted) so you're at the mercy of the real-world tolerance fit. This could be remedied with a spring (ala Noctigon KR1/KR4) or a design that doesn't necessitate cinching the tailcap all the way down (ala the old Surefire G2), but the former wasn't chosen and the latter would crash into customer expectations that the components should be screwed together completely (and likely PCB damage).
So my ... V2 I suppose since it has a captive switch ... either drew the short straw on tolerance stackup or the frequent
calibration damaged the PCB. The V1 is less temperamental but still randomly stops working on occasion and needs tightening.
Ergo, I'd suggest electronic lockout over mechanical for FWx flashlights.