U.S. Krags (all made at Springfield), unlike Norwegian Krag models, have clearance between the rear of the bolt guide-rib and the receiver.  (On Norwegian Krags, bolts were fitted to make contact at both the locking lug and guide-rib).  When they got into Civilian hands, some U.S. Krag bolts had the rear of the locking lug 'lapped' so that both the locking lug and guide-rib made contact with the receiver.  This was probably 'OK' and sound engineering, if a barrel was being fitted and adjustment was made for the added 'end-play' and increase of head space.  If no adjustment was made, the potential exists for excessive head space.
  I have only owned one Krag that had excessive head space.  Cartridges fired in that rifle showed obvious case stretching.  The bolt in that rifle appeared to have had metal removed from the rear of the locking lug.  The bolt's guide-rib made contact with the receiver.  The problem was solved by replacing the bolt.
  Today, I measured the guide-rib clearance on three model 1898 Krags.  The bolts were stripped and pulled rearward so that the rear of the locking lug made contact with its recess in the receiver.  A 'feeler' gauge was inserted between the guide rib and receiver.  Clearance varied from .003" to .004" on the three rifles.  I also examined a fourth Krag that was re-barreled by Sedgley.  The guide-rib made contact with the receiver for a 'double lock-up'.  However, the 1905-SA barrel, re-chambered for .30-40, was fitted for correct head space.
  If you encounter a U.S. Krag, and the (stripped) bolt guide-rib has no clearance with the receiver, it could be a Red Flag and may have excessive head space.

---

This 1898 Krag, re-barreled by Sedgley, appears to have clearance, but actually the lower part of guide-rib is in contact with receiver.

---

Some have theorized that the bolts which cracked were overhardened, and that those that weren't overhardened  will stretch, with the single locking lug acting as a lever, but not cracking, until the guide rib bears too.  That's someone else's theory, perhaps plausible.  I do know that lapping the locking lug to get the guide rib to bear can create excess headspace, depends on how far one has to go.  My Lt Col friend did that with a new bolt body, I have it here, and his rifle with my barrel has another new one he bought to correct the headspace problem he created.  Locking lugs must be softer than the receiver, he lapped it but most all material removed came from the locking lug.

If the difference between the original fit and that after lapping is only some .003"-.004" - that shouldn't make much practical difference with good brass and neck-sized handloads. Original headspace would have to be pretty loose for an .004" increase to put it past the practical "field" limit of .070" to .074" (depending on whose standards you prefer, US or UK). Heck, I've found more variation than .004" in the rim thickness of various .30-40 cases checked over the years. Personally, I'll continue neck-sizing and not worrying about headspace. My favorite quote from Gen. Hatcher:

"If there is excess headspace, the case will be stretched; if the headspace is too excessive, the case will be separated . . . where the thick brass in the base of the case begins to thin out as it joins the walls. When the gun is opened, the base of the cartridge will come out with the extractor, but the forward part of the case will remain wedged in the chamber and the gun cannot be used again until this is removed. . . With good strong actions and moderate pressure loads, and especially if shooting glasses are worn, this trouble is not particularly dangerous, but is annoying and inconvenient. Very little, if any, gas escapes to the rear, because the back end of the cartridge, which remains in the chamber, acts as a seal. In fact, it is just such short brass cups that the Germans have for years used as the breech seals in their heavy artillery . . .

"Our riflemen are used to measuring the headspace of their rifles by thousandths, and are prone to become very much alarmed if the headspace gauges two or three thousandths of an inch more than the normal maximum. This is a good safe attitude to take, but it is something like measuring cordwood with a micrometer. . ."

Julian Hatcher was a pretty savvy gun guy.  There is comfort in the 'gas seal' properties of rimmed cartridges like the .30-40 Krag and .303 British.  I'm a firm believer in neck sizing only in my bolt guns.

The separated "cup" does a fine job of sealing in rimless cases, too. I've done enough torture tests like the one illustrated below to know this effect rather intimately.

  (You need to Login   (You need to Login

The joy of having just one .308 and one 30/40 lies in having to neck size only.  I've yet to separate a case.  Neck sizing and decapping is done with a Lee collet die; that's all.  Simplicity.

Are you using the same collet die for both the 308 and 30/40?

You can neck size also by simply backing your full-length sizing die out one turn.  FWIW

You might try the Lee 303 British collet neck sizer die, for the 30-40. You could turn down the mandrel a thousandth or so if needed, to get a tighter bullet fit. I have had good luck with it.