I have read those papers but they have an essential problem.

They're wrong.

Or less succinctly, they fall prey to the "I can make shit up, too" fallacy, which to be fair is a very common state of affairs in the cryptocommodity space.

Proof of stake is not a distribution agnostic Oracle. Proof of stake doesn't actually prove anything as a second order function. It demonstrates what entities have the most interest in the further existence of the means of establishing their stake, that part's true, but it's not actually oracular unless you assume that the future will always be exactly the same as the past. The same concerns, the same resources, and just as importantly as every other element – the same players.

That's just not so and will never be so.

The whole pressure that proof of stake generates is toward centralization. It is against individuation. If the more resources you have, the more say you have and how resources are generated and distributed, and you actually care about how resources are generated and distributed (no matter how someone else may judge how/why you care), then you are going to act to get more of those resources. The very process of doing so means that consensus is less about population intent and more about your intent. That is the nature of the architecture.

In order to implement the consensus mechanism rooted in proof of stake you have to invalidate either the proof of stake or consensus. These things don't travel in the same basket. No amount of math, no amount of jiggling code around, will change that essential underlying fact.

The real solution is to move philosophically away from a top-down expectation of "oracular" wisdom and toward a system which builds individual representations and gateways and content from the individual out. To stop confusing the actions of crowds with the wisdom of crowds.

I don't expect to see that taking off around here anytime soon, unfortunately.