The same tension applies to a hypothetical mesh-network Internet, compared to the telco-backbone Internet: good for freedom and autonomy, bad for usability and performance. Alas, consumers tend to vote with their dollar for the latter. The "siren server" problem described by Jaron Lanier in "Who Owns The Future?" is a wicked problem when it comes to marketplace traction.

In the case of both mesh networks and dapps, I think the greatest potential is in the developing world, where centralized infrastructure is absent or inadequate.