Several organizations have launched websites where visitors post questions to be answered by politicians, and then a popular vote is used to choose the questions which are asked. All of these systems are basically scams: they're designed to fail, and they allow weak questions to rise to the top while tough questions languish.
For a tangible example of that, MoveOn used their membership list to vote a question for Barack Obama to the top of the list of possibilities, and it was asked of him in a televised debate. The incredibly weak question was about net neutrality, a topic he was already on record as supporting. In fact, a very basic search had his position on that topic as the first hit. In other words, it was a complete setup, a Soviet Union-style sham.
And, something similar happened with the same effort another time.
Now, of course, all of this depends on how you define "failure". If the goal is to have politicians asked tough questions or be confronted with their lies, these systems are complete failures. However, if the goal is to have politicians asked open-ended questions that they can answer with ease, then these systems work remarkably well.
In almost all cases, the questioners seem to have spent very little time anticipating likely responses, and as a result almost all of these questions can be answered with stock boilerplate. And, in almost all cases, the questioners don't seem to have researched the previous statements of those being questioned. Part of that may be a form of conditioning: most people have only seen the weak questions asked by TV newsanchors and the like and aren't familiar with tough questioning.
Make no mistake: politicians are quite happy with this setup. Since many of them are lawyers they have little trouble offering a vague answer that sounds like it answered the question. It provides a show, a verisimilitude of a debate. From the standpoint of many questioners, it offers an ego boost; perhaps those who offer weak questions should consider the ego boost of actually doing a public service.
The alternative to such systems is to have a set of known quantities (pundits, bloggers, etc.) vote on questions based on their toughness. And, all their votes will be public. Using that system, those known quantities will risk losing credibility if they vote up weak questions or vote down tough questions, they'll put their reputations at risk. For instance, a transportation policy expert will lose credibility if he votes down a tough question he disagrees with at the same time as he votes up an Obamagirl video.