Given one of your arguments is against flatline barrels, paintballs are an important example.
In any case, the factors you're quoting are overstated. People do go "It can't work! Look at that experiment with raw eggs and boiled eggs!", but that misses the fact that the only way the experiment can work at all is because the white and shell do impart force on each other. And paint is a lot more adherent and viscous than egg white.
The scale is also important. Torque increases in proportion to radius, but moment of inertia increases in proportion to the square of the radius. So the angular acceleration is inversely proportional to radius.
A small radius sphere filled with a viscous liquid actually approximates a solid sphere better than some critics would suspect.
Still, it's clear paintballs can pick up a random spin with ease, which translates into poor accuracy.
At worst, the spin induced by a flatline barrel has to be giving at least some degree of gyroscopic resistance to generating a random spin - after all, there's enough spin to capitalise on the Magnus effect.
And less random spin is a good thing.
Does that thing kinda look like a big cat to you?