This is an animation from the SOHO satellite of a comet hitting the sun on 12/24/1996. (To see a larger version of this, go to the offsite link http://seal.nascom.nasa.gov/gallery/LASCO/ or use, http://seal.nascom.nasa.gov/gallery/LASCO/xmas_c3.mpg. Warning 477K.) (LASCO is the Large Angle and Spectrometric Coronagraph on the SOHO satellite.)
The Ball-of-Light Particle Model predict that the impacts destabilize the core of the sun. For relatively small colliding objects, the cause and effect mechanism works like this:
In the press release that came out with this animation, scientists stated the solar mass ejection was unrelated to the comet's collisions. The Ball-of-Light Particle Model predicts the collisions caused the ejection.
(See also, Two Comets Hitting the Sun) (Large file! 245K QuickTime.mov)
This is not a coincidence! Three comets hitting the sun, three solar mass ejections, with the same timing between the comet's collision and the CME.
The Ball-of-Light Particle Model predicts that a comet doesn't even need to hit the sun to cause such a coronal mass ejection. If it is large enough, fast enough, and induces a large gravitational wave as it passes the sun at perihelion -- its closest point to the sun in its orbit -- then, the gravitational wave will induce electromagnetic waves across the sun's core which could induce a ball-of-light which could explode which could cause a coronal mass ejection. It is possible that such an event could even cause the sun to go Nova, or Supernova.)
(I believe that the exploding star Eta Carinae is an example of this.)
(See also, Gravitational Induction of an Electromagnetic Wave on a Ball-of-Light.)