I'm a member of ConsensusResearch group https://github.com/ConsensusResearch . We made some simulation tools around (Nxt-like) proof-of-stake consensus algo & published some papers around them.
One the possible improvements to proof-of-stake we're willing to investigate now is proof-of-stake + proof of-activity hybrid, where proof-of-activity is implemented like noted in PoW+PoA paper http://eprint.iacr.org/2014/452.pdf. From my personal point of view TenderMint is also about kinda PoS+PoA hybrid but implemented in another way. Do you Tendermint folks have simulation tools or may be even stricter models around your consensus algo to play with?
One the possible improvements to proof-of-stake we're willing to investigate now is proof-of-stake + proof of-activity hybrid, where proof-of-activity is implemented like noted in PoW+PoA paper http://eprint.iacr.org/2014/452.pdf. From my personal point of view TenderMint is also about kinda PoS+PoA hybrid but implemented in another way. Do you Tendermint folks have simulation tools or may be even stricter models around your consensus algo to play with?