As have others here. The granular structure at the breaks in those photos was commented on IIRC. I don't know if that was within the normal range or reflected progressive embrittlement due to stress.
The recoil lug in the body that was in closer contact with the bolt lug would have taken more load than normal obviously. Whether that would cause deformation of the lug in the body I don't know, but are they not surface hardened? Sub-surface deformation? Seems likely that the lugs would already have already "compressed" from prior loading as much as they ever would?
As for bolt "set-back" during proofing I will leave that to the experts to explain as well!
Another virtue of the design seems to be that if the bearing is unequal the worst that can happen is this kind of failure, rather than ejection of the bolt, because no matter what happens, that large recoil lug isn't coming off the bolt body.