If the target state is never achieved on one or more named systems, rhts-sync-block will halt the recipe until the local or external watchdog times out. It would be useful if a timeout option could be specified, such that rhts-sync-block would terminate with a zero error code if the desired state was achieved and a non-zero error code if the timeout was reached.
Turns out this was added last year, I was just updating a section of the docs that claimed it didn't exist.