Further Research Leads

The register report is the authoritative source for everything above. These come from FamilySearch's crowd-sourced Family Tree instead — unconfirmed, but worth chasing down.

1. Eleanor's surname. FamilySearch names her Eleanor "Nelly" Willis (1790–1840), matching the Rowan Co., NC marriage-bond scan already found (Barnet Wiatt & Nelly Willice, bond 20 Jun 1808) — but Pinkney was later born in neighboring Wilkes Co., not Rowan, so the link isn't airtight.
2. Five more children. FamilySearch credits Barnett & Eleanor with 10 children, not the register's 5 — adding William (1807–1863), Aaron (1815–1870), Joseph C. (1823–?), Jane (1825–?), and George Lee Davidson (1827–1891). This would explain why the 1830 census household already on file shows more children than the register names.
3. Two generations further back. FamilySearch extends the line past Barnett: father John Youngcastle Wyatt (1765–1843) & Hannah Park (1768–1862), and in turn John Wyatt III (6 May 1743 – 11 Nov 1815) of Rowan Co., NC & Rachel Vincent (married 26 Nov 1760, Ashe Co., NC). John Wyatt III's page cites a real, checkable source — "The Heritage of Rowan Co. North Carolina" by Ruth Beaver Wyatt, pg. 642 (Call # 975.671 H2h, V.1) — plus a will dated 5 Nov 1815 that (per FamilySearch's summary) never actually names his wife. The same page lists a sibling with an obviously wrong 109-year lifespan (1763–1872) and a "daughter" born 6 years before her parents' marriage, so treat the whole profile with real caution despite the book citation.