I think your theory is sound. By deferring pressure signs, as I said in another thread "sweeping pressure under the rug" you will compress them into less grains of powder. At the same time, when you get there, the size of the loading increment will give you very little gain. Unless you are a muppet and loading 5gr increments, as Gimp said, then everything should proceed as normal.
Now the point I made was that pressure and powder charge are not actually linked. So the issue here with the highest quality brass in individually machined actions is probably extremely trivial compared to people following the loading book with random brass and questionable rifle tolerances.
So you still have to stick to the one thing you have, which is deformation of the case. The case is the weak link. In 6.5x47L it is a very strong weak link. 6.5x47L is a case built from the ground up by wizards. 6.5 Creedmore is a wildcat given a thickened case base as an afterthought (by the same wizzards).
Mostly these things will be in the hands of people shooting them over chronographs in high end actions where you can get to the scary bit and back off.
Mine were doing 2800 with 140s (don't have the rest of the info on hand) and I have dug out 3 of my >300 3X fired brass and I don't have any case head deformation. As I mentioned in another thread I had ejector marks on one factory loading and completely flattened primers on another, when at the range the other day. Wincherster and Sellior and Belliot. Two different rifles. A Sako and a Ruger.
Bookmarks