If I remember Emary's article accurately, he tested loads in gain-twist barrels as well as the constant-twist M41 referenced on the data sheet. From an engineering perspective, the force required to enlarge the engraved area as a bullet progresses down a gain-twist barrel is a minor factor considering how little additional metal must be displaced and how much momentum the bullet has acquired before the twist sharpens significantly.
The pressure anomalies with reduced loads apparently occur as the bullet reaches the end of the fairly long freebore in the Carcano throat. When there is sufficient pressure and momentum at this point to fully engrave the bullet, regardless of twist pattern, the pressure curve is normal.