So, the bolt lugs either stay in front of the receiver lugs when fired, securing the bolt 100%, or they do not.
If the lugs are only partially engaged, and you managed to fire the gun, the contact surfaces will undergo more pressure, but that doesn’t mean the bolt moved back toward you any more than if they were fully engaged. It also does not mean there was more headspace.
If the bolt didn’t move backward forcefully and basically unrestrained, (unmistakable) then it didn’t move at all for all intents and purposes. In theory you might be able to depress tiny triangular indentations of extremely shallow depth into the receiver lugs if you didn’t have a ramp such that your bolt permitted pin fall with almost no closure.
If the loud, damaging round was fully contained, the resulting damage was not because the headspace grew larger than expected. Headspace is all the cartridge can know. How much or little lug contact exists is impossible to discern inside the chamber. This reasoning does put it all on the load or an obstruction.
Bad brass that expanded or ruptured or just plain excessive pressure with good brass, that did that to the brass could explain it.
Are you resizing brass by a very significant amount? There is probably some level of resizing that is so extreme, that the firing pin “will just barely” fire the round, but the headspace is truly too big for the chamber by any standard and the brass cannot handle the stretching permitted.