Gunplumber, your comment below is interesting:
Check out some brass at the range and see how the bulge usually is off to one side. It could be a series of alignment problems or just a large chamber with the shell off to one side when it goes bang.
That bulge is typically just in front of the extractor groove and it's as common as trees in a forest. Having cut the head off several cases so bulged and measured case wall thickness all the way around, one thing was easily learned. The thinnest part of the case wall bulges out. When case wall thickness is very uniform, there's typically no uneven bulge.
Virtually all cases get pushed against the chamber wall at their back end by the extractor. So the back of the cases are all off center in the chamber when the firing pin strikes their primer. (Note: there may be some straightening or better alignment of the back end of a bottleneck case to the chamber axis as its shoulder centers in the chamber shoulder from firing pin impact.) Extractors are spring loaded and press the case rim off dead center on the bolt face. Mauser claw style ones press the case head to the left in right-hand actions. Those sliding ones in the bolt head like Winchester's post-'64 push feeds push the case head up when the bolt's closed. In any instance, the head of the case gets pushed as far as the chamber or bolt face recess allows.
Indexing rounds in the chamber with a mark on them showing where the extractor pushes them, firing the round then noting where the bulge is shows interesting data. Careful inspection of new cases show where the extractor scared the case as it was chambered. And a magnifying glass reveals firing pin imprint details on the primer (especially if you file a flat on its round edge to leave a significant mark in the primer dimple); another way to see how the case was indexed in the chamber when it fired. The case bulge is seldom at the point where the extractor pushed the case head off the chamber centerline. Proof to me that that bulge is common only to the thinnest part of the case wall in front of the extractor rim.
Your other comment about case heads against bolt faces is good; few folks know about this:
When you neck size, what makes you think that shell is going back in the chamber the way it came out? A good bolt action will cam it in but the base of the case will not sit on the bolt face flat without twisting the case. A good indication is brass rubbed off on the face of the bolt or bolt lip. There is a lot more there than mentioned in the load books.
Few factory rifle bolt faces are square with the chamber axis. It's common for one point at its edge will be a few thousandths further forward than a point opposite it. Firing a brand spankin' new case with a perfectly square case head in such a rifle makes that case head flatten against the bolt face at the angle of the bolt face and pretty much stay there. Neither full length nor neck only sizing squares that unsquare case head back to square. When that round's chambered again and not indexed such that it matches the bolt face, there's two problems.
One is headspace; depending on how the case indexes against the bolt face, the clearance between the case and chamber lengthwise will vary a few to several thousandths. Stand a fired case on a flat surface against a V shaped object then rotate the case. If the head's out of square, the case mouth will enscribe a circle and not rotate about its center. The more the case mouth spins off its center, the more out of square the case head is.
The other is accuracy. Tests have shown that when the high point of the case head is aligned with the locking lugs on a squared up case head, the barrel whips such that its muzzle shoots bullets away further away from group center than when elsewhere. Square case heads shot against squared bolt faces shoot most accurate and stay square when sized for reloading. An unsquare case head against an unsquare bolt face is bad for best accuracy. Which explains why smart 'smiths square up bolt faces if best accuracy is important. Military rifle teams learned this decades ago when they tried to reload fired 7.62 NATO match ammo cases for their M14 and M1 service rifles. The best of those rifles would shoot all day long inside 4 inches at 600 yards with new commercial match ammo. Full length sizing and reloading those cases resulted in no better than about a foot at 600. None of the 'smiths rebuilding those rifles squared up their bolt faces.
One interesting thing Palma Team members learned about case head squareness. New ammo's typically the rule for international competition; everybody uses the same lot of ammo. Its case heads are fairly square but sometimes those at the limit of specs will be shot. The more lugs there are on the bolt, the more accurate those rounds will shoot. Which is why 3- or 4-lug actions are favored for this type of competition. A British riflesmith in the early 1970's, George Swenson, designed the first 4-lug action with very square up bolt faces to shoot their arsenal made 7.62 NATO ammo (the only type allowed in their fullbore long range matches) and it was an instant success. Since then, other makes have come on the market.