Reiterating all this…
So okay, it seems the driver board had something cooking on it, enough to discolor the board itself through-and-through, even on the battery-ring side.
The cells are reportedly okay.
The light still functions.
Right there, that rules out batteries and driver itself, and points to the charging circuitry.
If it wasn’t expecting, couldn’t tolerate, and yet was hit by, say, 20V, any protection circuitry (PTC resistor, TVS, etc.) could’ve started conducting and shunting away any excess, just like a 7135 that’s hit with too high a voltage.
So it cooks, and cooks, and cooks, and protects everything else, but starts heating up and frying the board. Sounds like it’s doing exactly what it’s supposed to do, but instead of protecting against a transient condition, was left to cook overnight.
People here probably remember my pushback against using usb-c in everything, and one reason was I never liked the idea of a 5V-only device being hit with 9V/12V/20V, even by “accident”. Well Hell’s Bells, what does this sound like is going on?
Only a proper post mortem will be able to figure out what component(s) cooked, and then find out what it is. Once that’s done, then you’ll know what the problem was.