fix(worker): 记住轮询间隔内的容量需求

异步视频任务在 River 轮询空档会释放本地 lease，采样瞬间可能观察不到已经触顶的领取需求，导致自适应容量长期停在初始值。本提交记录采样窗口内出现过的容量饱和，同时保持空闲 Worker 不盲目升容。\n\n移植时保留主线现有的单节点容量阶梯、请求下限和严格失败门禁，仅合入失败路径先记录并排空任务与回调、再删除模拟器资源的清理顺序，并补充脚本守卫。\n\n验证：go test ./... -count=1；go vet ./...；bash -n；ShellCheck；production-acceptance-script-test.sh。

scripts/cluster/run-production-acceptance.sh

@@ -381,12 +381,6 @@ cleanup() {
   if [[ $remote_load_drivers_installed == true ]]; then
     cleanup_remote_load_drivers >/dev/null 2>&1 || true
   fi
-  if [[ $single_node_emulator_installed == true ]]; then
-    cleanup_single_node_emulator >/dev/null 2>&1 || true
-  fi
-  if [[ $single_node_acceptance == true ]]; then
-    cleanup_single_node_kubernetes_resources >/dev/null 2>&1 || true
-  fi
   if (( status != 0 )) && [[ -n $run_id && $failure_recorded != true ]]; then
     set +e
     if [[ $single_node_acceptance == true ]]; then
@@ -398,6 +392,12 @@ cleanup() {
     mark_run_failed "${failure_reason:-acceptance workflow exited unexpectedly}"
     set -e
   fi
+  if [[ $single_node_emulator_installed == true ]]; then
+    cleanup_single_node_emulator >/dev/null 2>&1 || true
+  fi
+  if [[ $single_node_acceptance == true ]]; then
+    cleanup_single_node_kubernetes_resources >/dev/null 2>&1 || true
+  fi
   rm -rf -- "$temporary_root"
   exit "$status"
 }