From their description, RethinkDB don't have secondary indexing. So query will be very slow. Only key-value retrieval is useful. I find their query syntax too fancy without actually giving power.
I wonder what are the good point on RethinkDB?
In comparing to MongoDB.
JSON - BSON - limited serialization, slow MVVM - master/slave - better for performance, but still conflict will left unresolve and waste of storage for multiple version copy.