Remove point-in-time HOST2 status from the codebase

Comments asserting a partner is down, being rebuilt or not yet installed are true on the day they are written and wrong afterwards, and nothing prompts anyone to revisit them. Behaviour is described generically; current state belongs in operator memory, not in tracked files.

Plugin/unraid/api/rsync.php

@@ -150,7 +150,7 @@ $history  = [];
 // cleanup, docker updates — and run on their schedule whether or not RSYNC_ENABLED is true. So the
 // page could report a healthy 56-minute "daily sync" for a subsystem that had not run in a month,
 // which is exactly what it was doing: the last entry here is 2026-07-16, the day the global gate
-// was closed for the HOST2 rebuild.
+// was closed.
 //
 // Deliberately outside the 30-day cutoff. The last transfer is 29 days old as this is written and
 // would have dropped out of the window within days, taking the page from a wrong answer to no

Plugin/unraid/api/watchdog.php

@@ -35,7 +35,7 @@
 //
 //   Served from cache, collected only on a miss.
 //       vv_wd_all() SSHes to every configured partner, so its cost is set by the slowest node
-//       rather than by how much data there is. Measured at 8.3s on this host with HOST2 down —
+//       rather than by how much data there is. Measured at 8.3s on this host with the partner unreachable —
 //       paid by every visitor, every 30 seconds, because the tab polls. The cache check happens
 //       before the heavy include, so a hit costs one file read and no SSH at all.
 //