Docs still say QEP-3 is an open PR, and four other recorded facts have drifted

[mmcky]

A sweep of this repository's living documents and the migration tracker on 2026-09-02 found five facts that were true when written and are not now. None affects the published site. All affect what the next session believes before it starts work, which is the thing UPDATE.md exists to protect.

Where	Says	Now true
README.md § What this repo is; CLAUDE.md § What this is; UPDATE.md § Source of truth; ROADMAP.md § the type decision	QEP-3 is "still an open PR", so cite the type as proposed	QuantEcon/qeps#7 merged at 2026-09-02T01:15Z. The document on main still carries status: Draft, so the conclusion stands, but the stated reason is false, and a reader who follows the link sees a merged PR and cannot tell whether the guidance still holds
UPDATE.md § Known follow-ups	Bump actions/deploy-pages@v4 when convenient	Done. deploy.yml on main pins deploy-pages@v5, upload-pages-artifact@v5, setup-python@v7 and checkout@v7
UPDATE.md § Maintaining the contributions and feedback loop	contributions/ holds the source behind the four action-style-guide issues (#18 to #21)	contributions/issues/ holds seven bodies, 01 to 07; the text accounts for four
#30, first acceptance criterion	myst build succeeds with all 353 pages present	lectures/_toc.yml lists 358 pages: 357 file: entries plus the root. The difference is the five site-level pages outside the lecture reports and series indexes, so the criterion as written would pass a build that silently dropped all of them
#28, Background, final paragraph	"a 27-row ranked table per series"	The qe:series-ranked tables are 27, 52, 56, 68 and 145 rows; 27 is the smallest. The wording is inherited from REVIEW §6 in the design bundle, which is a dated snapshot and stays as it is

Proposed fix. One documentation PR for the four QEP-3 sentences and the two UPDATE.md items, saying "merged 2026-09-02, still status: Draft" so the guidance keeps its reason. Then two in-place body edits: the page count in #30 and the ranked-table wording in #28.

The design-record permalinks in #28 were corrected in place earlier today; they pointed at a pre-review commit of the bundle that said the trackers had not yet been filed.

Acceptance. Every sentence above matches live state, and #30's acceptance criterion counts every page the table of contents lists.

Found from QuantEcon/workspace-themes; the report theme that this migration depends on is tracked at QuantEcon/quantecon-theme-report.mystmd#2.