Refresh v5 migration guide performance numbers with a real benchmark run

[stijnvanhulle]

🎯 Changes

The migration guide's Performance section (docs/5.x/migration.md) had speed numbers with no reproducible source, and no memory data at all. Replaces them with the output of a real run of kubb-labs/kubb#3832's scripts/benchmark/v4-vs-v5 harness against the current v4 (4.39.2) and v5 beta (5.0.0-beta.104):

• Both time and memory, per spec and plugin combination, instead of time only.
• A link to the reproducible benchmark script instead of an unsourced claim.
• The <SpeedComparison /> chart embedded directly in the guide.
• A short explanation of why v5 memory usage is worse than v4 on the smallest spec (fixed adapter overhead that a 21-operation spec is too small to amortize) but pulls far ahead on larger specs, which the old table didn't cover at all.

Also updates blog/v5.md, whose "up to 6x faster" headline and OpenAI operation count (242) were stale relative to the new run (now "up to 5x", 281 operations, since the OpenAI spec is fetched live and has grown).

✅ Checklist

• I have followed the steps in the Contributing guide.
• I am using USA English (serialize, color, center, canceled, analyze, license).
• I ran the humanizer check on any new documentation I wrote.

---

Generated by Claude Code

[changeset-bot]

⚠️ No Changeset found

Latest commit: 2687412

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR