While a job queuing, such as ones formed on internal grid networks and supercomputers, would be unique and potentially beneficial for scientists, allowing users to select their own projects is a much more efficient and beneficial system.
In an arguably Capitalist v.s. Communist debate, the Capitalist ideology works better in volunteer grid computing. Just like in a Communist structure, job queuing works well when there is much overhead and analysis of submitted jobs.
However, the OCG is not built to be a multi-purpose grid computing platform. Allowing users to select which scientific project to join encourages competition. The OCG makes the assumption that users are generally smart. Users value their computing power and prefer to donate their computing power to causes they personally believe are important. It's up to the scientists and developers to convince users that their project to join their project, and it is this force that makes projects better. Without competition and choice, the OCG will not succeed.
Nevertheless, a paradoxical way of proving this philosophy wrong would be for a scientist to develop a project on the OCG that actually does multi-purpose experiments by randomly giving different jobs to different clients.