docs(audit): RAM/CPU placement audit + codify capacity goals

Add CAPACITY-GOALS.md (80% RAM / ~50% idle CPU targets + workload placement
model) and AUDIT-2026-07-29.md (live end-to-end analysis of all 7 hosts). Key
findings: memory goal met fleet-wide (highest 72%); CPU idles 0-6% (chase via
workloads, not hardware); tsys5 is single-CPU not dual; cnode4/5 are stale
running VMs. Answers the wnode-sizing question (grow tsys9, not tsys1), confirms
tsys6/7 for RackRental, and gives a DDR3 compatibility decision tree (RDIMM-only
servers reject consumer UDIMM). Refreshed STATUS/docmap and the returned audit
logs, and fixed an empty-array abort in deploy-check.sh.

💘 Generated with Crush

Assisted-by: Crush:glm-5.2

perf/deploy-check.sh

@@ -148,7 +148,11 @@ while [ "${#remaining[@]}" -gt 0 ]; do
             new_remaining+=("$host")
         fi
     done
-    remaining=("${new_remaining[@]:-}")
+    if [ "${#new_remaining[@]}" -gt 0 ]; then
+        remaining=("${new_remaining[@]}")
+    else
+        remaining=()
+    fi
     if [ "${#remaining[@]}" -gt 0 ]; then
         log "still running: ${remaining[*]} (${#remaining[@]} hosts)"
     fi