That is the correct process; if you want dedup&compression on all he disks in the disk group.
You missed step 0 though...
Have current and tested backups of all critical information; then:
- place the host in maintenance mode with 'evacuate all data to the other hosts' (long process...),
- delete and recreate disk groups one by one, (ensure that dedup and compression is enabled on the disk groups).
- exit maintenance mode & proactive re-balance it.
If you miss the re-balance that is ok, when you put the next host in to Maintenance mode (assuming you are adding a disk to all hosts for a homogenous environment) it will take advantage of the space on the empty host. The pro active rebalance should only necessary after completing work on the last host.