I wonder what kind of role it plays in your vision of a smart contract platform built on top of HAF. Is it a blocker, or a nice-to-see change?

It's not a blocker per se as the smart contract system would work without it, but it is very important, because it will reduce the CPU and IO resources required to execute the contracts (because there will be fewer of the more-expense reversible records to process).

Is the code public already? I couldn't find it in the hive repo.

No, it's still sitting in a dev's local repo right now. I wanted testing to be completed on all the other locking and P2P changes we're making before introducing a new variable, just so it is easier for us to isolate the source of any problem we find and be sure that we're testing OBI on a solid base. All the tests we're developing for those previous changes will also be very useful for testing OBI, because they generally revolve around creating forks.