Note that voting in general are several subtopics, and people mess them up. Even maintenance messages are a kind of voting, even if only one person leave the vote. I believe maintenance messages are an especially good example of why we need microvoting.
Assume that one person put a maintenance message on an article about spell-checking. The vote will then be whatever he put there. The next one to inspect the page should then be able to cast a vote on the spell-checking dimension. When the concensus reach a certain level, then the article can be put in a category for articles with dismissable spell-check templates.
I wonder if there should be one maintenance template, and that spell-checking is one dimension. Another dimension is verifiability, and a third is formatting. That is the template sint "spell-checking", it is "maintenance".
Another example is cases where you pose a question to the community. It should be possible to ask the community to chose on scale along a dimension, and the answer should neatly accumulate as part of the answer together with some confidence level.
A third example could be giving credits to a discussion post. That should be fairly easy to implement on Flow. In this case it is important to add some cost. One simple option is to simply limit the total number of credits a person can give or take, thereby making it important for people to use the credits wisely. This is the Slashdot model.
I have a paper somewhere about the importance to connect such in-site voting with some cost. At wikipedia the cost could be accumulated through some typing measure, but it is also possible to assign credits on random. Both works. It is also important to check how well voting from any user match up with a conrol group, if there are to much deviation the users opinions should be ranked lower. This is a bit dangerous, as it is important to keep deviant voices, but not let them be destructive.