Deciding Between a Local Review and Ultrareview

You're the sole maintainer of an open-source repo, on a Pro plan, about to merge a PR that migrates the auth token refresh logic. You've already run /code-review max locally and fixed the two issues it found. A contributor suggests: "just run ultrareview before merging too, it's basically free."

1. Is "basically free" an accurate description? What actually determines whether this specific run is free, and what happens once it isn't?
2. Mechanically — not just "it's more thorough" — what would running /code-review ultra on this PR add beyond what /code-review max already gave you?
3. Propose a one-line policy for when this maintainer should reach for ultrareview versus treating a local max-effort pass as sufficient, given the limited free allotment on a Pro plan.