Separate Populations?

What are my neighbors doing? Curiosity about that question resulted in some conflicting data. Ordered by when I added the RSS feed for them.

  1. search.twitter.com for “Athens GA”  – results are full of people talking about Athens, GA not in Athens, GA. Useful for people coming into town for an event.
  2. TweetLocal search for “Athens, GA” (or 30605 get same results) within 20 miles – Over the last 24 hours the RSS feed has given me 12 posts. First 5 users in search before 9pm: JeremyAce4 in Athens, GA, justdandelions in athens, ga, bozaf in Néa Smírni, Europe/Athens, aaronbarton in Athens, GA, elbee103 in Athens, GA (last @ 7pm). The hit on Europe/Athens is pretty disappointing.
  3. search.twitter.com for “near:AHN within:20mi” (or 30605 or AthensGA get same results) – Over the same 24 hour period, its RSS feed has given me 53 posts. First 5 users in search before 9pm: ThePicMan, julieteaston, ryan_lafountain, RyanHague, alester (last @ 7pm)
No overlap. How is that possible when they supposedly are coming from the same population (time, space, and active)? Both services look for their data on Twitter. Both are looking at the self-identified location for Twitter users. Both have the same range. So, why do they have such different results?
Looking specifically for the Tweetlocal users in search.twitter.com reveals them in the results. Searching on a user though doesn’t reveal the location. On the profile is the right location, so they should have been in both results.
Both fail in my opinion.