We've identified the only remaining issue with random API latency issues and errors and are deploying a fix.
We are currently monitoring and taking action as needed to speed up recovery while everyone reconnects.