Some background to the driver problems, and how DEL improved it with a simple mod.:
Nevertheless these FETs were poor, even with the gate driven with a full cell (4.2V minus Schottky diode drop, and sag/internal resistance) after DEL’s mod. At real world cell voltages they are far higher resistance.
See DEL’s graphs at GB BLF LuckySun D80-SE Need long tubes? -3/3/16 - #2417 by DEL and GB BLF LuckySun D80-SE Need long tubes? -3/3/16 - #2426 by DEL
As supplied, getting only half cell voltage, minus Schottky drop, with say a cell at 3.7 or even 3.5V, sagging further under load, they were barely operating. That’s ignoring the poor tail spring used.
For comparison, a decent FET as used in a good BLF FET+1 is in the 5 milliohm class, at all cell voltages.
I suspect that prototype evaluation was only ever done with full cells. Also by others here. Not realistic.
Then there was the suspicion that the prototypes supplied to Dale etc. had superior emitters, production units from lower bins. He could never reproduce the results with production units. Even then, they never took the current that would be expected with a good FET.
The PWM was about 150 Hz. Horrid. And maxed. at 98% on turbo, yet another loss.
Then there was the nonsense with the tailcap switch retaining ring. Originally it had a shoulder on it to contact the tube, but they were sloppy, forever coming loose, and even failing to hold together at-all. My early one suffered this. So instead of doing it properly and going to a LH thread and making the threads fit properly they removed the shoulder and cranked them in super-tight, nearly to the point of stripping. But in removing the shoulder they forgot to lengthen the tube to compensate. Hence all the short-tube problems, and the need to make longer replacement ones, at extra cost. Famously Kronological “totally missed this”.
I just glued mine together with threadlock, which probably makes the switch unmaintainable.
Like I said, attention to detail is everything, as the FW3A team have found. Sometimes it seems like one step forward, two back.
If Lucky Sun can step their game, this would be a nice refresh. The original team sometimes expressed frustration at communication difficulties and un-discussed changes, for what should have been a straightforward tweak of an existing product, basically LED, MCPCB, driver UI changes, new logo, job done.
For this to be good, I see it needing a lot more work, but worthwhile.