The way it is now I couldn't tell you right away what version supports what just by glancing at the docs for a minute.
I recently ran into issues with replicaSet myself; finding it hard to locate documentation on using user-based authentication. It boiled down to me eventually locating the necessary info on the Master-Slave page for ensuring I did a db.addUser() on the slave's local db. All in all, I'm much looking forward to a rewrite of the documentation.
I believe CouchDB is a better choice for very large data sets because of its design.
+ CouchDB uses a Map Reduce design that I believe would scale better over very large data sets. + CouchDB always stores data in a consistent state on disk. You can literally pull the plug on the server at any time and the data will never be inconsistent.
MongoDB is geared for performance and is a great bridge between a relational database and a high-performance No-SQL database. But I don't recall that it's strength is handling large datasets.
the theory behind the thing is great. in reality, its buggy and not fun to work with.