explorer: extract controllers from zoomWatcher + build-step decision (deferred from #249) #303

Follow-up to #249: deferred Explorer controller extraction

#249 ("should we refactor explorer.qmd before the next big feature?") is closed — the agreed first tranche shipped: CI Playwright smoke gate, characterization tests, and extraction of pure functions into ES modules (sql-builders.js, explorer-utils.js) with unit tests (#287). The URL/mode-state refactor (#208) then completed on top of it (#213/#288/#289/#301).

This issue tracks the deferred portion of #249's original roadmap so it isn't lost:

• Split controllers out of the large zoomWatcher / mode-transition block — extract the imperative map/selection logic into named controller modules (continuing the strangler pattern, not a rewrite).
• Decide on a build step (Vite/TypeScript) for the extracted modules — explicitly punted during Architecture: should we refactor explorer.qmd before the next big feature? (analysis + options) #249; revisit now that several pure modules exist.
• Consolidate selection mutations into a controller (overlaps with explorer: consolidate selection mutations into a controller (deferred Step 2 of #187) #189, deferred Step 2 of explorer: extract _selGen freshness primitive (post-mortem on PR #186 6-round Codex loop) #187).

Context: surfaced in a 2026-06-19 audit + Codex fact-check review. The point: #249 being closed reflects that the agreed chunk is done, not the entire original wish-list. No urgency — this is "clear before the next big feature," consistent with #249's framing.

Related: #208 (done), #249 (done), #189, #187.