The fine thread on the tail does not help with the loosening of the RH threaded switch PCB retaining ring, too easy to apply a great deal of pressure/friction at this interface. A coarser square-cut thread would be great. An LH threaded ring is also desirable, so that it tightens instead of loosening when the tail cap is removed.
This problem can happen straight away, or after a period of actual use. Never knowing whether the tail assembly is going to come apart, or just come loose and intermittent, each time you change the cell is not confidence-inspiring. My bodge was to thread-lock it together.
Masses of discussion on the original thread, maybe read it completely, in case there are some lessons to be learned there that may have been forgotten or never addressed.
Not bashing LS, but I have struggled with the non-standard driver cavity on mine. having had to file down an adapter PCB, and the driver, just to fit a 17mm one to replace the original. Tedious. Just saying.
One of these:
https://www.fasttech.com/products/1138202
Alternative would have been to solder copper wire around the edge of the driver to expand it, but that would have been mechanically and electrically poor.
Also be aware that the driver cavity is fully anodised, in fact everything is, except for the faces of the tube. There is no current path to the driver or tube, it is reliant on contact through the driver retaining ring. No problem, it works well as long as the retaining ring makes a good contact with the driver ground ring. But there are quite a few such critical interfaces in series in this torch, relying on perfect contact at each.
Other torches do things differently, e.g. masking off anodising to ensure solid contact through the tail/tube/head for the important -ve cell current path.