I appreciate the distinction that Daniel is trying to make between what he calls "partitioned consensus" databases and "unified consensus" databases, but I don't know if the points that he makes about partitioned consensus truly generalize to all such systems. In the context of this blog post, he specifically calls out partitioned consensus databases for requiring two wide-area round trips in order to run 2PC. Howeve…
My take away is that it depends on your requirements. If most of your transactions are multi-entity, then Calvin-like is better. If most of your transactions are single-entity, then Spanner-like is better.
I implemented projects in GAE Datastore, though it's classic Paxos inside, it's clearly multi-partition database. For my workflow almost all of the frequqnt transactions were single-entity, so multi-partition Datastore worked fine with it.