Intro- Yes, this has been discussed before, but I have a different solution and some updated reasoning. Please, for the love, read the post before down voting it... (but then you can down vote it all you want, I know it's not a popular idea :)
I personally think users should be able to flag answers that have technical inaccuracies (controversial as that might be). Then we should have expert reviews of flagged answers, and if the answer is incorrect it should be noted at the top of the answer by the expert reviewer so regular users can easily see the contradicting argument and vote the answer down if they feel so inclined.
Why? Well it has to do with how SE is changing. Back in the good old days of SO, most of the people who were active on the site were technically competent individuals who worked as programmers and sys admins. They had direct experience with the questions at hand, and could upvote because they had used the answer before and it worked for them.
Fastforward to today.
The site I'm mostly active in is Aviation.SE (though there are dozens of very similar sites at this point), where you have a core of knowledgable individuals who have worked in the Aviation Industry and then you have hundreds of people who are "enthusiasts", who range from private pilots all the way down to people with only simulator experience (yeah, that'd me in that last group there.) So a lot of the people voting on that site may not have direct experience with many of the questions being asked.
But, I'll be darned if it stops any of them from voting.
This is how "urban legends" are started, right? A group of semi-knowledgeable people get together and come up with a reasonable explanation for a phenomena, they all nod their heads a lot and, rather democratically, decide they must know how it works.
Of course the problem with this is that physical laws don't really care what you think. If one person knows the right answer, and everyone else believes a different answer, you're going to get that wrong answer voted up and the right answer will be snubbed.
This is a huge problem facing SE sites these days. Let's face it, back in the SO days you could just pop open a terminal (generally) and figure out "real quick" if an answer was true or not. But the ability to explain the answer to "How does a wing generate lift" can take years of schooling to get a grip on.
Solution 1: I think that in communities where "group thinking" may end up giving us the wrong answer, there should be an ability for people to ask for an "expert review". Over on Aviation.SE, most of the regulars trust Peter Kampf or fooot (as examples) to know the physics behind how a plane actually works. It would be awesome if we could flag an answer for their review, so that their sizable knowledge can be used to see through the mists of "common knowledge".
If one of the experts (who probably ought to be elected) decides the answer is wrong then they would flag it as such, along with a brief explanation as to why. Preferably near the top of the answer, or at least in a place where it's easy for people who are reading the answer to notice.
Now, I like democracy, so I personally think that at this point we could just leave it be and let people vote the answer down themselves.
Solution 2: I just go make my own site if this isn't what the community wants ;). I mean, honestly, isn't that what one should do at that point?
As a final note: Some of you may say "well, that's what the second and third answers are for, aren't they?" Well, no. Two reasons. Firstly, answers are not supposed to be replys to other questions. Secondly, well, it goes back to group think. If the first answer is highly popular (even if it's wrong) and it sounds reasonable, people will often vote it up without even reading the second response...
Anyway. This whole thing is open for discussion, I'm just tossing the idea out there is all.