interesting.
I would think that the cases that would be good candidates for being improved would be ones that have a significant taper, low angle shoulder, and long necks.
cases like the 30-30, 22 hornet, 30-40Krag.
more modern cases like the 270 the gain would be minimal.
I won't disagree with you on the smaller cases. Straightening out the taper, increasing shoulder angle and decreasing neck lengths can increase powder capacity by as much as 15-20% with such small cases.
The 30-40 Krag has a very long neck and long taper on both the body and the shoulder, again, capacity increase is significant. Same with the 30-30 but not as significant.
I did a 338-06 improved and was actually surprised at how well it did, when compared to the standard 338-06.
The body had a one degree taper, shoulder 45 degree taper and the neck length was reduced to .250 in. The volocity increase was appx 150fps but the recoil from a 24 inch barreled Ruger No1 was more than I could handle, without developing a flinch.
I cut that barrel back on the tenon end so that the chamber would accomodate a 308 Win length case.
Did the same thing with the dies and mounted the barrel onto a 98 Mauser action.
It's a 338-08 Improved now.
Even with the one degree taper on the body, 45 degree taper on the shoulder and short .250 neck, I can't get more than about 50 fps velocity increas over factory 338 Federal offerings.
I've used 338 Federal brass to fireform in my chamber, which it does very nicely, but even when I load to maximum for my rifle pressures not a significant difference.
Very accurate rifle.
Parker Ackley's original experiments with improving already existing cartridge cases weren't as drastic as the later versions. His original intent was to stabilize harmoncs by creating a condition in the case with the sharp angle shoulders that would allow a more efficient/consistent burn rate with the off the shelf cannister grade powders available to him at the time. Velocity increase was appreciated but not the initial intent.
As you've likely noticed and suggested, some of those improvements gave significant increases in velocity, but mostly, they seem to have better predictable and repeatable accuracy, which, IMHO, is really why so many people like them so much.