A recently savaged meta post (Disable privileges if a user has been inactive) seems to have been inspired by the OP's bewilderment that so many votes were cast (and not for him) in an election for DIY.stackexchange.com... He was trying to figure out why so many people participated in such a low-activity exchange, chalking it up to long-inactive users who suddenly showed up for the election. In the course of the conversation, someone pointed out that
Users who are eligible to vote in an election receive a notification in their global inbox letting them know it's voting time
Given that eligibility to vote is 150 rep points, and that you're automatically awarded 100 points when you join a new stack, it's likely that many - if not most - eligible voters on low-activity stacks may be only occasional participants (answered or asked one or two questions)... this may weaken the voting power of regular and active community members, and may tilt elections towards candidates with more recognizably from other exchanges (SO, etc) even if they haven't been particularly active in the exchange for which they are candidates.
So 3 dependent questions:
Do people feel that this is occurring (what do the actual voter stats tell us? Are elections in low-activity sites dominated by relatively inactive members?)
If it is happening, is it an issue? Maybe it's good to favor candidates with broader SE presence
If it is happening, and it is an issue, what's a solution? I know that any suggestion of weighting votes will send the meta crowd into a down-voting frenzy, but what about making the amount or rep required to vote in low-use exchanges higher (though this might make the voting community too small)? What about not broadcasting a specific election to eligible voters who have low-rep in that exchange AND haven't visited the site in
X months? They can still vote if they show up, but they're not necessarily encouraged to do so.