I think the idea Steve's putting out there is that P2P applications over secured channels are inherently hard to block (deep packet inspection can do this, but it's rather fragile). And that implementing them in Perl is a possible idea.
His idea is not meant to have a business model. it is meant to be computer-mediated, networked communication, not dependent on central servers, solely for the purpose of communicating, and for communicating in situations where the "norma" channels are suppressed. It is not meant to make any money, and its developers will, no doubt, either not care about being paid, or will feel that striking a blow for free communication is payment enough. It is meant to be a means of responding to the Bloggers and Twitters responding to pressure to censor themselves -- by censoring themselves and pretending that they aren't.
It's BitTorrent for radical ideas, with Tor on top. Anonynimization plus decentralized communication. I think Perl probably is a good choice.