As of the creation of this subverse (and often before, and probably often always), Voat is in trouble. Servers are expensive, bandwidth is expensive, and while it's getting cheaper, Voat keeps growing.
One method used for fixing this sort of scaling problem is P2P, peer-to-peer networks. P2P has become a buzzword lately, with some successful implementations and some not-so-successful. By having everyone using a service share a portion of the load, the need for central service is lessened or eliminated, greatly reducing expense. P2P networks without a central server can also be very difficult to censor or shut down, a boon for a service like Voat.
That having been said, P2P is also fraught with issues. When messages propagate through a P2P network it becomes impossible to track down exactly where on the net they sourced from, meaning spam and brigading may become difficult problems. The security model for the network also needs to be properly designed, or channels could be taken over, or in a worst-case scenario, computers connected to the network could be compromised.
This subverse is for mapping out a protocol which would implement a multi-layer forum system, with up- and down-voting, with strong authentication for multiple levels of administration and for users, and possibly also encryption.
Note: Any code this project produces will be open-source under the MIT license.