After doing some research now I'm not so sure.. NAD83 supposedly would only cause a 1 or 2 meter different between WGS84 so that's probably not the problem. The problem could be their maps aren't precise enough for this-- after all their maximum zoom is not really that small if you try it. I can still see about a 3-block radius on their tighest zoom, which means maybe they stopped paying for really precise maps?? I dunno- maybe someone should try to update the python script to make sure i'm not missing something. My code should be fairly easy to translate to python I hope.