[FEATURE]: concurrent-ensure benchmark for missing-service reconnect herd #38805

Feature hasn't been suggested before.

• I have verified this feature I'm about to request hasn't been suggested before.

Describe the enhancement you want to request

Parent: #36285

Before landing reconnect-herd policy, we need a small missing-service reproducer that measures:

• contender launch count when N clients call Service.ensure against an absent registration
• recovery latency until clients are healthy

Existing coverage checks jitter bounds and concurrent reuse of an already-healthy endpoint. It does not measure the missing-service race.

No reconnect policy in the first cut — baseline current Service.ensure behavior so later PRs can keep only changes that reduce launches without delaying single-contender recovery.

Related closed draft ticket: #38569
PR: #38568

Environment

Same class of setup as #36285 (V2 managed service, multiple TUIs, macOS / next channel). This request is for a harness in packages/client, not a new end-user incident report.