Evidence #3:
It is an Ethereum public good
It is an Ethereum public good first because it incentivizes development of new oracles. Anything that strongly incentivizes development of Ethereum oracles is obviously and doubtlessly an Ethereum public good. The linked document should not describe how prediction markets work, because there are already published articles on this topic, I refer to them. I clearly described some potential use cases: salary to scientists and salary to free software authors (clearly there can be found more use cases, e.g. salary to journalists, support of NGOs, etc.) It does very concretely state what will actually be built: It is given a mathematical model to be implemented in the blockchain.
My work is not to create a new smart contract handling prediction markets, but use prediction markets for Ethereum public good (incentivizing creating oracles), possibly using an existing prediction markets smart contracts library like Gnosis or Augur.
In short: UNDOUBTFULLY an Ethereum public good, the challenge is wrong in every point.