The results vary with the bullet. In the '60s, A. A. Abbatiello measured concentricity on samples from, IIRC, 42 lots of National Match ammo and got 1 moa wider groups from 0.004" of tip-tilt off-center. Beyond that much tilt, additional tilt made little additional difference, as the bore straightened out anything much more than that. This was 30-06 ammo using the 173-grain M1 Type bullet. When Harold Vaughn did it with stubby 6 mm bullets, the same amount of tilt gave only about half the dispersion.
The differences include how many calibers the bullet bearing surface is (longer is straightened out more by the bore) and how far forward of the geometric center of the bearing surface (where the tilt is centered) the center of mass of the bullet lies. The further forward (or behind) the center of mass, the bigger the orbit it has around the bore line for a given degree of in-bore tilt, so the greater is orbital speed. When the bullet exits the bore the bullet is tossed at that speed radially away from the mean trajectory path in a direction tangent to the bore where it is nearest to the center of mass at the moment of release. The drift due to that tangential toss was on the order of half a foot per second with the National Match ammunition, and that is too slow for drag to have much effect on during bullet time of flight, and thus it stays with the bullet all the way to the target.