Caleb Phillips and I performed an independent assessment of MetroFi coverage. The standard set out in the contract between the City and MetroFi was 90% of outdoor areas within the proof-of-concept network. The definition of the proof-of-concept that metrofi and the city agreed to was within 500 feet of an access point. Turns out that by that definition, the proof-of-concept network area was only 1.6 square miles, less than the minimum of 2 square miles called for in the contract, and less expansive than the 2.5 square miles claimed by MetroFi in public.In our investigation, we made a random selection of locations that were outdoors within 500 feet of an access point. We went to each of 53 locations and measured the ability to make a connection (minimum 64kbps up/down). We found that we were able to make a connection in 31 of those locations and failed in 22. The probability of a network with 90% outdoor coverage having that ratio of successes and failures is 2 in a billion.
I believe that the fundamental problem is that at greater ranges, the access point is not able to hear the relatively weak radios in typical client devices.