docs(02-11): four-way re-measure, final per-surface verdicts, second Discover cause

Deployed the poller fix to archi-dev-box and re-ran the frozen harness
(02-PERF-FINAL.json, 5 runs/surface). Web5 fixed (275ms, below both
its 566ms pre-phase-2 baseline and the 300ms pass bar). Server's
regression closed (574ms, below 738ms baseline) though not yet under
300ms. Fleet substantially improved (790ms, down from a 2631ms
regression). AppDetails restored to at/near its own baseline.

Investigated the two open items the coordinator raised:
- OpenWrtGateway: this run's 5/5 samples failed with a Chromium
  "Target crashed" error cascading from an unrelated surface earlier
  in the same harness run — recorded as not-measurable, not written
  in as data. Separately confirmed via source (OpenWrtGateway.vue's
  h1 renders unconditionally, and a "No router configured" RPC error
  deterministically shows a real Connect-to-Router form) that the
  prior baseline/after/remeasure numbers were measuring a genuine,
  substantive disconnected-state UI render, not an empty/error page —
  so the "six confirmed regressions" count is not retracted, but the
  numbers are flagged as reflecting one specific code branch.
- Discover (1389ms, worst remaining, least improved): profiled
  directly and found a SECOND, distinct, phase-2-class cause —
  showStagger/card-stagger entrance-animation classes are baked into
  the DOM at first mount and never programmatically removed (the flag
  is a correctly-scoped once-per-session const, but nothing ever
  re-renders to strip the class), so every KeepAlive detach/reattach
  cycle restarts the CSS animation on reactivation, replaying the full
  entrance cascade on every revisit. Confirmed via a diagnostic
  showing DOM card count doubling transiently on every revisit and an
  extended animationstart/animationend event log. Not fixed this pass
  — the safe fix's blast radius spans 5+ files outside this plan's
  scope (Apps.vue, Marketplace.vue, Home.vue, several Web5 sub-cards)
  and needs its own real-device verification budget, matching the
  precedent 02-02's original KeepAlive rollout needed for this exact
  class of change. Named and evidenced, recommended as a dedicated
  follow-up rather than expanded into this plan under time pressure.

REQUIREMENTS.md's PERF-02/PERF-03 rows updated to the final state.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

.planning/REQUIREMENTS.md

@@ -120,8 +120,8 @@ Which phases cover which requirements. Updated during roadmap creation.
 | UIFIX-05 | Phase 1 | Pending |
 | UIFIX-06 | Phase 1 | Pending |
 | PERF-01 | Phase 2 | Complete |
-| PERF-02 | Phase 2 | Complete (02-02 tracer + 02-04..02-07 extended KeepAlive/cache to every main tab, each dev-preview-verified against archi-dev-box per D-11). Timing-regression verdict closed by 02-10 (`02-FINDINGS.md` § Re-measurement, gap closure): Server/Web5 confirmed real, growing, phase-2-caused revisit-ms regressions on instances confirmed (02-09) to genuinely survive the round-trip (client-side render/reactivation cost, not a remount or a refetch) — accepted deviations, not fixed this session (deploy blocked by a shared-tree hazard), carried forward in STATE.md Blockers/Concerns |
-| PERF-03 | Phase 2 | Complete. Timing-regression verdict closed by 02-10 (`02-FINDINGS.md` § Re-measurement, gap closure): Discover/AppDetails/OpenWrtGateway confirmed real, growing, phase-2-caused revisit-ms regressions (same client-side split-signal cause) — accepted deviations, not fixed this session; Wallet/send-flow's apparent regression cleared as environmental noise (re-measure returned at/below baseline) |
+| PERF-02 | Phase 2 | Complete. 02-11 (`02-FINDINGS.md` § Client-Side Render Cost Root Cause + § Task 3) named and fixed the real cause of Web5/Server's revisit-ms regressions — three leaked background pollers (`useFleetData.ts`, `FipsNetworkCard.vue`, `Web5Monitoring.vue`) armed in `onMounted` and never disarmed once their owning views joined `KEEP_ALIVE_PATHS`, gated to activate/deactivate. Web5 now fixed (275ms, below both its 566ms pre-phase-2 baseline and the 300ms pass bar); Server's regression is closed (574ms, below its 738ms baseline) though not yet under the 300ms stretch target — residual named as real, un-eliminated per-resource reactivation cost, not a new defect |
+| PERF-03 | Phase 2 | Complete. 02-11 fixed Fleet's leaked `useFleetData.ts` poll (790ms, down from a 2631ms regression, substantially closing the gap to its 330ms baseline). AppDetails restored to at/near its own baseline (1231ms vs. 1204ms) — residual is the already-documented `useCachedResource` per-mount setup cost, not fixed further. Discover (1389ms) has a SECOND, distinct, evidenced cause found this session (CSS entrance-animation replay on KeepAlive reactivation, `card-stagger`/`showStagger` never removed from the DOM) — named with full profiling/diagnostic evidence but NOT fixed (blast radius spans 5+ files outside this plan's scope, needs its own real-device verification budget) — recommended as a dedicated follow-up. OpenWrtGateway: not measurable this pass (Chromium crash cascading from an unrelated surface); prior numbers stand, confirmed to reflect a real (not empty) disconnected-device UI render, not retracted |
 | MNODE-01 | Phase 3 | Pending |
 | MNODE-02 | Phase 3 | Pending |
 | MNODE-03 | Phase 3 | Pending |