docs: select FTW delivery order and simple setup scope

[frahlg]

FTW now has an owner-selected delivery order: close the confirmed control and Lua-host gaps, finish daily charging, then simplify commissioning and make live evidence useful to people and agents.

Record the source/test/browser baseline in the maintained roadmap, including the nine baseline regressions from #1170/#1199 (seven failed, two passed on merged Core f1a3b76). Keep existing capabilities, missing flows and missing physical evidence distinct.

The owner has removed the STRÅNG / roof geometry / panel drawing / related settings stack from Core scope. A future optional extension remains undecided; normal setup must work without irradiance-source selection, azimuth or panel layouts. VISION.md states that boundary. Close the old Python-optimizer implementation work while preserving useful requirements for current Energyplan evaluation and household-goal reporting. Independent benchmarks and development tools remain valid.

Validation: reviewed current code, relevant PR diffs and reviews, ran the recorded baseline regressions through a temporary overlay, and rendered the webapp simulator during the baseline audit. Documentation changes pass git diff --check and the planning-document gate. No runtime code or worker artifact changes. These are selected priorities and evidence requirements, not claims of physical commissioning or a new release.

---

Note

Low Risk
Documentation-only updates to vision and roadmap; no changes to dispatch, APIs, or shipped behavior.

Overview
This PR records owner-selected product scope and priorities in planning docs only—no runtime, worker, or client code changes.

VISION.md now states explicitly that STRÅNG, roof geometry, and panel drawing are outside Core scope, with normal setup required to work without irradiance-source choice, azimuth, or panel layout; a future optional extension remains undecided.

docs/roadmap.md gains a 12 September 2026 implementation review section that: pulls the STRÅNG/settings stack and retired Python-optimizer PRs out of the Core queue while keeping useful Energyplan requirements; pins a reviewed baseline (Core/webapp/native/drivers/website SHAs) and documents seven failing / two passing regression cases from #1170/#1199 on that baseline; lists confirmed legacy control gaps (Lua driver_command, dispatch load/SoC, curtail freshness); adds an area-by-area “what exists vs what remains” table; and sets a numbered delivery order (control gaps → daily charging → commissioning → live evidence → external authority/savings → later thermal).

^{Reviewed by Cursor Bugbot for commit 79671c6. Bugbot is set up for automated code reviews on this repo. Configure here.}

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_4db26679-76db-4162-80c9-ffe66b61cd18)

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-12T10:02:49.449413Z	79671c6	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.