I think you're right that the Minimum Viable Product needs to "fleshed out" and concretezised a bit. But my idea so far is to try to run one or two projects with the blog / email as only supporting infrastructure, because it's fun and to learn about the problem "first hand" so to speak.