I will state first thing, this is JMHO and I have no data to back this up:
I think it depends on the drillings used previously. On slightly older balls, every hole going into it will effect the core, some slightly, some more. If drill #2 was significantly different than drill #1, you now have changed the core dynamics, even if it is only slightly, 2 different ways. Now if you plug and redrill it a 3rd way, totally different from drill #1 and drill #2, now you have again changed the core dynamics a 3rd way. Even if all are just slight changes, I can't see how a ball will roll true having the core dynamics changes 3 different ways. I don't think that is something any manufacturer could predict or would take the time/energy/resources to model. If all 3 drillings are relatively close, I would think you might be ok, but it is still probably a crapshoot as to whether the ball will roll true or not.
As far as rolling over the plugs, I was always told to try and drill a ball to avoid that. While the new plug material may be urethane, it won't match the cover of the ball, in most cases, and you would have a small area where the ball will not have the same traction as the rest. In the overall scheme of things, it may not make a difference, but if you a buying a ball to use on heavy oil, for example, that one spot may make a difference. It would be like trying to drive in snow with a bald spot on every tire.