I assume that by using a straight edge to test headspace, you are saying that you placed a loaded round into the chamber and then used a straight edge to verify that the rear of the case was flush with the barrel. However, a loaded round might experience resistance to chambering all the way which might give the erroneous impression that the head of the case remained flush because the rim was engaging the bottom the the rim recess. By the same token, a fired case might have to be forced back into the chamber and would show the same resistance to chambering as a loaded round.

Actually, headspace in a rimmed cartridge is defined as the distance from the bottom of the rim recess to the face of the bolt, block, or action. Even if the mechanical headspace is correct for the rifle, use of a cartridge with an unusually thin rim will result in excessive actual headspace. When the firing pin falls on such a cartridge, the entire cartridge is forced forward until it meets resistance in the form of the bottom of the rim recess. When the powder charge ignites, the walls of the cartridge obturate and cling to the walls of the chamber, while the head of the case is forced rearward by the build up in pressure, causing the case walls directly in front of the solid web of the case, which does not obturate, to stretch, eventually after repeated firings resulting in a rupture.

The same phenomenon occurs with belted cases with excessive headspace, which have not been headspaced on the shoulder, rather than on the rim. It can also occur with a rimless cartridge, where the rifle's headspace is correct, but that of the loaded cartridges is insufficient.

I suggest pasting progressively thicker bits of paper to the bottom of resized cases, then trimming the paper to the exact shape of the cartridge rim and attempting to close the action on the chambered cartridges. Merely attempting to close the action on a sheet of paper is irrelevant.
Better still, make or obtain a proper set of headspace gauges and use them.