Start small, think big, move fast… Why not start with a group dedicated to individual attributes? Define those parameters. Establish the data structure. Define the tools to get to, update, and terminate that data. Then move out from there. Imagine a world where companies or individuals can buy any IT product off the shelf-without the current bureaucracy associated with major purchases-whenever they want, with whatever bells and whistles they want, because the data it uses is universally structured.
It would save billions of dollars (after the Great Data Structure in the Sky was set up anyway). Imagine a world where data is separate from the IT driving it. You can have a database but you can’t change the structure or meaning. Some elements may be un-updateable except by an authorized group (e.g., source data like date of birth) and a master version captured at whatever level deemed appropriate-maybe national then synchronized with a world-wide repository.