chore(rfc-0014): sweep 12 'future PR-B' comments when WithMetrics bridge lands

[trilamsr]

Context

Post-wave-audit finding #7 (see docs/v1-rc1-post-wave-audit.md).

10 "future PR-B" comments in module/ + 2 in docs/ reference the RFC-0014 WithMetrics bridge that has not shipped. They are factually accurate today, but become stale the day PR-B lands.

Locations:

• module/processor/patterndetectorprocessor/patterndetector.go:357,366,447,642
• module/processor/patterndetectorprocessor/pcie_aer_test.go:47
• module/processor/patterndetectorprocessor/thermal_throttle_test.go:252
• module/pkg/patterns/thermal_throttle.go:42
• module/pkg/patterns/cuda_oom.go:80
• module/pkg/patterns/pcie_aer.go:70
• docs/patterns/pattern-4-thermal-throttle.md:145
• docs/integrations/prometheus-scrape.md:381

Fix

When RFC-0014 PR-B lands, sweep all 12 sites in the same PR (or as the first follow-up).

This issue is the tracker, not the fix.

Acceptance

• All 12 comments either deleted or rewritten to reflect post-PR-B reality.
• Closes with PR-B's merge commit.

[trilamsr]

Deferring this issue from the post-wave-audit cleanup bundle (PR sweeping #383 / #381 / #379).

Precondition check: RFC-0014 PR-B (patterndetectorprocessor.WithMetrics bridge) has not landed — git log --all --grep=WithMetrics returns only wave-3 port commits, none of them the bridge. docs/rfcs/0014-metrics-to-logs-pattern-input.md:3 still lists status as "accepted, blocks issue #260 PR-B", and docs/integrations/prometheus-scrape.md:381 still describes PR-B as the future-emit side.

All 12 "future PR-B" comments are therefore still factually accurate as of 2026-06-01. Sweeping prematurely would either delete load-bearing forward-looking documentation, or replace it with stale post-PR-B prose that lies about the codebase state.

Keeping this issue open as the tracker. It closes with PR-B's merge commit per its own acceptance criteria.