In brief
The recent source reviews changed how Ancient Measures chooses its next piece of work. An existing article is a proposal to examine, not a conclusion to defend. Before adding detail, the review asks whether its central question is useful, distinct and supported by sources. It then records what the sources establish and what the site's own analysis adds.
Start with the question the evidence can answer
The earlier workflow could make an inherited outline look more settled than it was. A title, a place in the queue and a list of desired books could become reasons to keep pursuing the same argument. That is a poor use of research time if a narrower comparison would tell the reader something better.
The revised review considers the article alongside nearby pages. It can retain the question, refocus it, replace the treatment, merge overlapping material or retire an unhelpful page. The choice needs a recorded reason and recoverable earlier text. An unavailable book remains an acquisition gap; it does not by itself prove that the subject has no support.
This changes the order of work. Evidence requirements follow the claims the revised article actually retains. A source requirement attached to an abandoned argument cannot silently become a permanent obstacle to a different, supportable investigation.
Make the original contribution checkable
A claim ledger separates reported measurements and historical statements from calculations, comparisons and interpretations made here. Original arithmetic does not need a borrowed citation pretending somebody else performed it. Its inputs do need sources, and the calculation must expose its units, assumptions and limits.
The Stonehenge review provides a concrete example of the resulting work. Its release corrected a disputed annotation, retained the approved figure's geometry and added a test readers could inspect. The question was not answered by polishing the old paragraph. The argument, source treatment and explanatory tools had to be reviewed together.
The same distinction matters in the methodology: a close numerical match can be a result of a calculation without establishing an ancient designer's intention. The public explanation needs to show where that inference would begin.
Keep design acceptance separate from historical support
Eleven approved temple infographics now have a preservation check. Their composition and geometry remain protected while documented factual corrections receive their own evidence trail. Approval of a figure's design does not certify every surrounding historical claim.
Likewise, a publication count does not report how many historical claims have been rechecked. Narrative assessment, source-based scholarship, editorial enrichment and unit-graphic acceptance retain separate measures on Analysis. A build-log entry does not advance those measures.
The practical result is a more demanding choice before drafting: explain the question, the evidence and the contribution first. Extend an article when that strengthens its argument, rather than because an old outline still has empty sections.