There's an interesting conversation in the comments at VentureBeat on a post written about Idearc's partnership with AmericanTowns.
The chicken and egg conversation revolves around whether "content" must be deep or "geography" must be wide... in order to reach a critical mass of "local" visitors on a hyperlocal site. In english, does a site need to cover a lot of towns (eg topix, AmericanTowns, YellowPages) OR does a site need to have a lot of high quality content for many individual towns (topix, citysearch, local newspaper websites if they were banded together with other town newspapers)?
I don't think they are mutually exclusive. But, it raises a good point. There are very few websites that do both. Topix might be the exception of a site that covers a lot of towns and has a lot of great content [and community] for each town. But, all they cover is news and commentary.
There are a lot of content types and verticals left that don't have deep content and wide geographic coverage, such as events and classifieds.
It is interesting that topix has a thriving community. I imagine it is because they were able to aggregate a lot of great local content for a lot of towns. The content came first. Then, the community. I'd challenge that that is the necessary order of things, though, based on the fact that social networking websites were able to spread virally from a few different locations. It'd certainly be interesting to see the adoption rates of popular social networking sites as a function of time and geography.
Is there a hyperlocal killer application out there that will propogate content + community at the same time? Kinda like craigslist, but with less control over the geographic rollout. Maybe like eVite did, but faster, and with more useful data than home bbq listings.