rationalization / post-hoc. If you were setting out to make the next MySpace or Craigslist, would you consciously decide to make it ugly as a competitive advantage? Ugly designs happen by accident [re: neglect] and in spite of the underlying site's success. By the time the site is successful it's too late/unnecessary to change it.
I disagree. As other's are pointing out, a site's design says a message, and having an ugly design might send the right message.
Dove is using this idea in their "Campaign for Natural Beauty" (http://www.dove.us/#/cfrb/) . The models aren't ugly, but they also aren't the usual perfect-looking models.
When I was in grade school, I remember there were anti-smoking and tobacco posters on the wall of severely aged, sick looking people along with some text about how "chewing cost me my mandible". Those were examples of actual human ugliness, and they worked.
So that's all to say, ugly could be part of your message. Which isn't really a novel idea.
What's more interesting, in my opinion, is when something seems bad from a usability standpoint and still succeeds. plentyoffish.com fits under this category.
