safekeeper: check for non-consecutive writes in safekeeper.rs

[arssher]

wal_storage.rs already checks this, but since this is a quite legit scenario check it at safekeeper.rs (consensus level) as well.

ref #8212

This is a take 2; previous PR #8640 had been reverted because interplay with another change broke test_last_log_term_switch.

[github-actions]

3808 tests run: 3702 passed, 0 failed, 106 skipped (full report)

---

Flaky tests (4)

Postgres 16

• test_lfc_resize: debug-x86-64

Postgres 15

• test_delete_timeline_exercise_crash_safety_failpoints[Check.RETRY_WITHOUT_RESTART-timeline-delete-before-schedule]: release-arm64

Postgres 14

• test_location_conf_churn[1]: release-x86-64
• test_pull_timeline_partial_segment_integrity: release-arm64

Code coverage* (full report)

• functions: 32.5% (7423 of 22847 functions)
• lines: 50.6% (60114 of 118755 lines)

* collected from Rust tests only

---

_{The comment gets automatically updated with the latest test results
3427fd1 at 2024-09-03T15:02:07.914Z :recycle:}

[problame]

A new unit test to assert we hit precisely the added error in the presumed scenario would be great.