The mesh system has the following capabilities:
This system should be quite scalable. Both qTox and IPFS are based upon distributed hash tables (DHT) so that each peer does not need to store the full index of data for the entire network. Gossiping between SSB peers may be slower, but the small world effect will presumably still make for quite efficient delivery in a large network. Caching or pinning of IPFS data and its content addressability means that if a file or blog becomes popular then performance should improve as the number of downloads increases, which is the opposite of the client/server paradigm.