There is a beauty in rimmed cartridge cases like the .30-40 Krag and .303 British.  They Head-Space on the rim thickness (not from the bolt face to the chamber shoulder).  In the event of a case separation, the case head and rim still keep most of the hot gases in the chamber and barrel.  In a rimless cartridge, there is not this 'extra seal' and a large volume of hot expanding gas comes harmfully into the Action.  I do not take head separations lightly.  The .30-40 and .303 British should be viewed as 'Head-Spacing' both on the rim and on the chamber shoulder.  Every time these cartridges are fired, the case shoulder and body move, if necessary, to make contact with the the chamber walls and shoulder.  This is why it is desirable to neck-size only, when possible.  From my experience, the U.S Springfield Armory seems to have held pretty tight tolerances on Krag chambers.  All of my 1898 Krags will accept, or almost accept, cases fired in one of their 'brothers'.  The front locking lug and metallurgy of the Krag are not as "springy and elastic" as I have found (rear-locking) Lee-Enfields to be.  I believe the British also utilized greater tolerances in their chamber dimensions.  This is maybe good for service in the field but bad for re-loaders!  I once had a Short Magazine Lee-Enfield that was re-built in 1953 and had a brand new B.S.A. barrel on it.  I reloaded for it with new commercial brass and a Lee Loader that neck sized only.  By the third loading, I was getting complete head separations.  Not a whiff of gas to the rear and good accuracy on paper, but that darn rifle destroyed my 303 Brass base!  I have not encountered head separations in my Krags.  I have had an occasional case split near the shoulder in Remington Factory Ammo (probably 1 time in every 75 rounds).  I feel confident with Krag safety but annealing case necks, neck sizing only, and "feeling case interiors" for incipient case separations is good practice.