I assume you are using Banzhaf Indexes (If I spelt that right!) to determine the "swing vote" power.
I have long thought there is potential for individual investors in say pension funds to be allowed to vote their 'beneficial' (#) holding their way - so the controversial measures start to become more democratic. One day I will persuade a pension fund to try this, but that might just play havoc with your analysis :-)
Out of interest, how much of say the Dow have shareholders that hold 'disproportionate' power ? _ say 10% more influence than their vote count would suggest?
Good luck !
(#) I don't think the term I mean exists.
edit: * -> # to avoid italics