docs: address review feedback on `prev_blockhash` validation

docs: address review feedback on `prev_blockhash` validation

- Clarify `CheckPoint::insert` eviction semantics when `data.prev_blockhash`
  conflicts with the checkpoint at `height - 1`.
- Explain placeholder handling in `merge_chains` conflict branch where
  `u.data()` may legitimately insert `None`.
- Preserve mismatch height in `merge_chains`' fallback so release-mode
  callers get a useful `try_include_height` instead of `0`.
- Reframe the "update displaces invalid block" test comment to make the
  best-effort recovery intent explicit.
- Derive `PartialEq` on `CheckPointEntry` and fix a broken intra-doc link.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>