After reading the paper, I see many problems but the advantages seem to be giving agency to projects being built on top of steemit, by specializing the communities and interaction with already existing APIs that benefit Steemit without integrating them directly into the platform.

Suppose a person has Ph.D. and has no skin in the platform. Just by verifying their credentials in the real world, one could reward a type of reputation that allows them to contribute questions or to curate questions. Even if he's on steemit or not.

Four main advantages:

1. The current user interface of steemit is not ideal to present questions (something you are experimenting on, with the AskSteemit initiative)
2. A voluntary leveling system, based on merit, separated from the social aspect but funded by it.
3. Dapp interaction with other API's be it blockchains or not.
4. Allowing users to explore other lines of interests, without them spreading too much in their profiles.