That still doesn't make a lick of sense to me! It appears the LH lug fracture passes through the entire structure, either originating or terminating at it's outboard side, rather than it's root. There ought to be no load concentration there. IF it's the fracture terminus, then it's at least slightly conceivable. But then the question arises- From whence and when did the failure originate? Peter Laidler's comments seem worth investigating, if possible.
Brian Dick's example, on the other hand, is quite "textbook". Very nice! No fault of the design, just the operator/reloader.