I like the direction you’re going with routing and thinking about “bandwidth” between cities. Analyzing a graph of connections between users in different groups is the approach BrightID uses to identify sybils. Take a look at these papers on graph-based anti-sybil systems and see if any of them inspire you:
Actually, I wasn’t referring to BIO, because BIO is just an asset that has value based on its inherent properties. It wouldn’t be a native token and wouldn’t need to be used to pay for any services. Once you have BIO, the only thing you can do is trade it as a collectors item that originated from someone making a one-time contribution to BrightID.
It seems like it would be hard to enforce that our own deeply-integrated token would be the one a consumer of BrightID uses to incentivize users. Is there another use for the token besides incentives?
In your description of the treasury model, you said there needs to be a native token. We would need to find the purpose for the native token (see my other response: https://medium.com/@adamstallard/to-have-a-native-token-it-has-to-be-used-to-pay-for-something-986d98335358)