O.K. . . . not to get in to a big debate or duel . . . but when I read this originally and posted . . . I neglected to ask what Hawg and several others have already mentioned. I have a machining background and am still scratching my head . . .
If it was a "bored through" cylinder (i.e. as for cartridge) I can see how the chamber could be machined so that the mouth of the chamber (end towards the barrel) could be tapered to a smaller I.D. - a fairly simple mating process with the right tools. But we're talking a BP cylinder - bored from the mouth end (barrel end). How on earth can you machine it so that the mouth is a smaller I.D. than the lower portion (nipple end) of the chamber? I just don't see how it can be done unless you bored the chamber and then went in with an expandable hone of some sort that expanded after inserting. I really don't think they go to that extent.
So . . . I'm not casting doubts on what the OP has stated or questioning him . . . and yes, it's possible that his end measurements may be off due to the difficulty in taking them . . . but that still doesn't explain the balls moving forward. If he's getting a good shear of lead . . . they shouldn't be doing it. Or am I missing something?
I've never shot a "49 Pocket (although I'd love to have one). I'm assuming you are using about 12 grains of BP. Doesn't seem to me like that would force the balls in the unfired chambers forward unless there was a problem. Is it possible that the chambers are "out of round" - i.e. enough that you might get a shear and it look O.K. but off enough to allow movement? Just asking . . .