Add DuckLake golden perf corpus

[bill-ph]

Summary

• replace the targeted scenario's raw-view-only perf catalog with one golden corpus and one result artifact
• run four equivalent raw-Parquet-view / DuckLake-table query pairs: total events, one-day filtered event count, event-name aggregation, and distinct persons
• keep three DuckLake-table-only regression probes for persons totals/daily series and event daily series
• preserve the existing one warmup / three measurement-iteration cadence

Why

The scenario already creates and validates posthog.events and posthog.persons. The perf corpus should use those tables while retaining raw-Parquet controls for like-for-like comparison. The one-day table count is the direct perf probe for filtered metadata aggregation.

Validation

go test ./tests/mw-dev/scenario ./tests/mw-dev/scenario/perf ./tests/perf/core
git diff --check origin/main

[github-actions]

Test Impact Plan

Deterministic summary of how this PR changes tests, CI runners, and coverage-risk signals.

Summary

Area	Added	Changed	Deleted
Test files	1	1	0
E2E/journey files	0	0	0
Workflow files	0	0	0

Signals

• Test cases: +0 / -0
• Assertions: +0 / -0
• Skips or known failures added: 0
• Workflow continue-on-error added: 0
• Workflow path filters added: 0
• Test commands removed from justfile: 0
• E2E/journey retry lines added: 0

Coverage risk: neutral or increased

No coverage-reduction warnings detected.