I think it is Lyman that gives an average change in velocity of 10 fps per inch change in barrel length. There was no qualifiers on that number as to whether that was a minimum or maximum or an average number, the type of load or any other information.
I did run into a firearms forensics article regarding barrel length and buckshot where the velocity difference was minimal from 30" until one hit the 20" length. From that point, velocity dropped at a higher rate. From what I recall, the average change in velocity per inch of barrel change would have been in the 10 fps range though the change in velocity between popular hunting length barrels (24"-30") was negligible from a practical standpoint.
The biggest change in muzzle velocity readings occurred when velocity readings were changed from 3' past the muzzle to the muzzle. European loads were looked on with amazement when they became common on our shores 30+ years ago because they boasted such high velocities. I wouldn't be surprised if allowed operating pressures were a little higher as the majority of the Euro shells seemed cleaner burning than U.S. loads. Measuring them with the same process as American made shells of the time gave a different measurement and the differences in measuring practices became apparent. I'm guessing this is a C.l.P. vs SAAMI thing as this seems pretty common in other areas too.
There have been advancements in powder technology that have led to higher velocities but changes in measuring techniques still play a role.
