The diagram shows two blockstores – it's not clear that this is just one optional aspect that the diagram aims to share.

Under Advanced Deployment, we say

An Ops Manager deployment with high availability and data redundancy includes two servers that host the Ops Manager Application and Ops Manager Application databases, four servers that host two Backup deployments, and an additional server to host the arbiters for each replica set.

Deploy an HTTP Load Balancer to balance the HTTP traffic for the MMS HTTP Service and Backup service. Ops Manager does not supply an HTTP Load Balancer: you must deploy and configure it yourself.

All of the software services need to be able to communicate with the Ops Manager Application databases, and the Backup Blockstore databases. Configure your firewalls to allow traffic between these servers on the appropriate ports.

However, the Advanced deployment actually emphasizes two distinct (and not tied together) ways you can run Ops Manager:
1) You can run Ops Manager's application in a highly available fashion behind a load balancer (doing so may not be required, since if your Ops Manager application goes down you can manually start a new application if you have some downtime tolerance for Ops Manager's functions), assuming you also have a 3-data-node replica set for the Application database – which is required for HA of the backing db.

2) You can have more than 1 backup blockstore. In other words you can scale out the blockstores manually to have different backups associated with them. <--- this is important for folks with larger #s of deployments being backed up, but certainly may not be relevant for all users.

I think we're trying to show the Advanced Deployment to highlight all the different things you can do – depending on your needs. But we need to emphasize that there are permutations not shown that are simpler.