assert_ready: SHC readiness probes only the first inventory member; one down member blocks all deploys

[dtwersky]

roles/splunk/tasks/splunk_shc/assert_ready.yml queries the shcluster status endpoint on groups[target_shc_group_name] | first. That endpoint reports the whole cluster, but if that specific member is down/unreachable the uri task itself fails, blocking all SHC app deploys and removals despite a healthy captain and quorum. Routine single-member maintenance therefore blocks deployment.

Fix: iterate members until one answers (loop with rescue/until, or a failed_when that only fails after all members are tried).

Pre-existing from master; affects both the deploy and removal SHC gates.