The one thing that hasn’t been mentioned as a possible root cause of this problem is the alignment of the little-end bush relative to the big-end.
In my opinion it is essential that the little-end bushes be bored to size using the big-end of the connecting rod as an alignment datum. This of course means that such work has to be done before the crankshaft is assembled and also requires access to a milling machine and some careful setting up.
I did notice a reference to reaming the little-end which did make me cringe a bit! It is worth noting that as small an alignment error as a ¼ thou. that is 0.00025” (0,0064mm) in the bush, which is less than 1/10th the thickness of a human hair by the way, will tip a Vincent piston by 0.0008” (0,02mm) over its diameter.
If that misalignment happens to be on the axis of the gudgeon pin and is also quite possibly more than the above figures then it is easy to see that the compromised skirt clearances will result in extra friction leading ultimately to this sort of disaster. Such misalignments will do no favours to the big-end bearings either!