There was a good Twitter thread today about finality, explaining that a protocol can't guarantee both immediate settlement and liveness. Yet Ethereum found a compromise by running two different confirmation rules at the same time after moving to proof of stake. When a significant part of the network disconnects, blocks will still be produced but finality will be postponed. (OK, I see something like that is also in the OBI proposal, but it's unclear to me what will happen when no quorum is reached at all.)
It's not reassuring to read that @blocktrades didn't study either Tendermint or Ethereum's Gasper. If Hive makes this change, the chain might become more likely to stop working during serious disruptions, like Solana.
I wouldn't be too concerned about my research level of Tindermint. I reviewed it and fully understand it since I received complaints that I hadn't bothered.But as I suspected, it was a waste of time from my perspective. Tindermint has limitations and differences that don't make it a good fit for Hive. I'll post about it eventually, when I get past more pressing issues.
As to OBI's design, it's generally much more resilient to loss of BPs than Tindermint and can continue working down to a much lower number of reachable BPs.