Clustered Indexer Stats

Now that we have a supported set of indexer stats and expose them in the UI, it would very helpful to roll them into the existing REST API endpoints for bucket statistics in ns_server.  This provides clustered, per-node, per-bucket statistics for the whole cluster from a single endpoint on each node.  Without this, an administrator and monitoring system needs to be constantly aware of the cluster topology and gather stats from each indexer node individually...making it very hard to work with automated monitoring systems.