Make weekly forecasts more reliable and explain their basis

[TheBeems]

Weekly forecasts could extrapolate a short burst across the rest of the week, count sparsely observed cycles as useful history, and disagree with weekly warning notifications. This change makes estimates conditional on sufficient observations and explains the evidence behind them.

• Require 30 minutes of continuous weekly measurements and use at most six recent hours. Weight representative completed weeks by coverage and age, reducing the influence of recent bursts further into the forecast.
• Show observation duration, usable history and remaining daily/hourly allowance. Use appropriate date precision and break observed chart lines at measurement gaps.
• Use the same adaptive setting and calculation for weekly alerts and the dashboard. Document the behavior and add deterministic regressions and chronological synthetic benchmarks.

Validation:

• Native ARM64 suite: 370 passed, 0 failed, 0 skipped.
• ARM64 and x64 application builds passed; x64 build reported 0 warnings and 0 errors.
• Synthetic rolling forecasts reduced mean absolute error versus recent pace at 6 hours (0.527 vs 1.100 percentage points), 24 hours (0.631 vs 2.880), and reset (0.637 vs 10.733). This repeated-pattern benchmark does not establish real-user accuracy.
• Git diff whitespace check passed.

Pending verification: x64 tests need the .NET 10 x64 runtime. Integration preflights passed artifact and manifest checks, but overall preflight remains failing because the selected Debug build is not registered (the production Store ARM64 package is installed). Live UI/COM checks of the changed build remain pending in an isolated installation; the Store installation was not replaced.

[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 2026-09-15T08:58:34.024670Z	5a45e04	PR opened

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.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 5a45e04232

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".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".