The short answer is that obfsproxy is indeed not designed to facilitate PTs like DNS transports. It will probably require a considerable refactor of obfsproxy to write a DNS pluggable transport. It will probably require so much refactoring that I'm not sure if it would make sense to merge the changes back upstream (because the changes might not be useful for other PTs).
Re the above from George, and some talk I see about obfs versions, am I right that obfsproxy is not yet a general pluggable transport tool? As in a library and/or standalone proxy that two endpoints put in front of themselves, choose from a menu of (or negotiate) obfus and say go in a nearly automatic fashion? A C/python library interface would be interesting, as would deciding what the standalone proxy might be among socks5, transparent firewall redirection, tun interface, etc. In addition to DNS, people have written HTTP, IRC, ICMP, raw IP, etc tunneling tools in the past.