The ideal or what I call the ideal is with time lock and puzzle mining if you, me, with a team of others have a discussion on Tauchain which solves a problem in the world but we want our solution to remain private along with our knowledge bases for only a fixed period of time (say the time we expect to live (+60 years) then time lock encryption would allow us to make the promise in the smart contract form that all of our private knowledge base will be released after +60 years time no matter what.
This would allow the community to trust us because future generations would benefit not just from our solution but the knowledge we arrived at would not be at risk of being lost. The library of Alexandria problem is averted.
Mine to unlock would be to have a time lock puzzle which with ordinary hardware would for example take +60 years to unlock but with specialized hardware which would be some kind of ASIC then this could perhaps be reduced to 10 years, or 6 years, or 10 months, depending on how much computation resources society chooses to throw at it. In this way we can have a means in an emergency to release knowledge from a time lock but at a predictably high cost. The cost could be predicted for example as the amount of memory required or some other resource which in theory could be paid for perhaps using Agoras itself to rent these resources.
Mine to unlock is speculative because I don't see anyone actually having done it. But I don't see anything theoretically which would make it impossible to do either. So I think that is something which will happen sooner or later.