I manage several SQL server availability groups using Netbackup but do not use this technique. As we use FULL, Differential and Transaction Log backups we have to backup the primary replicas. I back up each node individually with a BCH file which excludes all databases which are READ ONLY mode as these are the secondary replicas. The system databases are always available for backup on every node. I have written a PowerShell script which interrogates the SQL instances and writes a BCH file which works for the node with its current Availability Group configuration. This is set to backup ALL databases but then has additional exclude lines for all the read only database. I get this to run by using an event based trigger which is initiated whenever a failover event is logged in the windows Event Log. I also get the same event to initiate a FULL backup after the BCH file has been created.
N.B. I do have a SQL instance with 4 nodes running multiple instances and 100+ databases and multiple Availability groups. There are frequent transaction log backup in addition to the full and differential backups. In this situation, when trying to perform a restore, using the GUI, we found that using Netbackup is impractical. If you have a environment like this I suggest you extensively test to ensure it performs as you would want. This testing should be done after you have a few weeks of backup history as it appear the processing of significant backup records slows down the GUI.