Homer Simpson's "bowling ball with a liquid centre" (from a catalogue he was reading once) would probably be more hype than help. A liquid centre in a bowling ball would tend to retard the rolling motion of the ball, decreasing its power and accuracy simultaneously. If it only operated side to side, it would mean the ball would tend to stay on track, but since the rotating motion is more down the lane than side to side, extra energy would need to be imparted to the ball to set up a liquid gyroscope effect.
I don't know enough fluid dynamics physics to back this up with solid calculations, but I did spend a good ten minutes rolling my water bottle up and down my desk. It didn't work very well.
Mokalus of Borg
PS - Also, my water bottle is not spherical.
PPS - Nor is it as heavy as a bowling ball.