I'm skeptical that's true in medicine, though it depends on how else they filter their community. My own experience is that non-doctors with strong opinions about medicine that they feel motivated to share are usually precisely the people you don't want medical advice from. Often they are adherents of one or another alternative-medicine system, or otherwise have really idiosyncratic views.
It's amazing how much wasted energy is spent on pseudo science. From zero-energy devices to "secret" cancer cures. Those involved create boogie men(usually the government) who are repressing "the truth", as the actual data to support their claims is weak. Rallying people against a common enemy is easier than coming up with an actual product that works. It's incredibly lazy, and usually a few minutes of research will cast doubt on claims. The scamsters are scum, but you can see the motivation. Those who actually believe are scarier.
The startup that immediately came to mind here is CrowdMed: http://readwrite.com/2013/05/08/crowdmed-wants-to-crowdsourc... . Crowdsourced diagnoses are a regulatory and liability minefield, but the upside to patients is huge and (in my opinion) worth the risk - instead of an impersonal WebMD, you get individual attention by experts to specific cases, with oversight from other community experts in the form of powerful downvotes.
N.B. CrowdMed's spend-reputation-to-answer "prediction market" methodology (in contrast to the free-to-answer StackOverflow/Quora model), described a bit in the article, might be of interest to other HN'ers. I would expect it to drive away potential contributors, but in the case of CrowdMed, it seems that the potential contributors with other incentives to contribute are also the higher-quality contributors (namely people who want to do good with their answers and are confident in their expertise) so it tends to weed out only the bad actors. It's a dynamic that's worth considering for anyone starting a Q&A site of any kind.