In "Weaving the Authoritarian Web: Liberalization, Bureaucratization, and the Internet in Non-Democratic Regimes," Boas, details primarily Saudi and Chinese control on the Internet. This really made me want to buy some land on the Bahrain/Saudi border and install a little headless Linux box connected to a high speed or multi-modem dialup unfiltered connection in Bahrain on one end, and a wifi cantenna rig (O'Reilly recommends Pringle's) on the other.
Actually, in for a dime, in for a dollar, might as well use the winning methodology of the most recent DefCon "security" conference's wifi shootout (the 2005 winner achieved 125 miles), or the 2003 winner, which managed over 30 miles with equipment (beyond the wifi card and laptop) under $100.
The underlying point to the Boas paper of course is that they don't have to be perfect, just good enough to make the few who can evade the restrictions politically insignificant. This will continue to be somewhat of a cat-and-mouse game, as costs for access to alternate technologies/networks will likely fall over time (e.g. satellite, cell networks from neighboring countries, mesh wifi networks that route out through foreign connections, etc.), not to mention brave souls willing to take risks, as Jake mentioned talking about the Chinese reporter who posted a "blistering letter on the newspaper's computer system attacking the Communist Party's propaganda czars and a plan by the editor in chief to dock reporters' pay if their stories upset party officials." (The plan got dropped, but the hero of the story eventually got fired and the entire section of the newspaper has since been shut down).
Regardless, these ingenious little (architectual) tools and original-sense hackers provide an invaluable resource to ICT development; the Pringle's Cantenna has already found a use for an Egyptian entrepreneur to connect his home to his Internet cafe, and I can only imagine there are other similar projects.
Creativity becomes almost as valuable as access in rolling out ICT infrastructure projects, and the same forces are at work -- laws (protecting monopolies or restricting radio frequency usage, for example), cultural norms (a mesh network requires cooperation and a method to arrange antennas to maintain a mesh and not get stolen), market (cost-to-connect, cost of equipment...), and architecture; but it is with the architecture that creativity can have the impact. The others (law, culture, market) are beyond the control in almost every development project's time-frame and budget, but the architectural challenges might be already being pursued by the "mice" of the world.