Send automated weekly summary report to site owners

## Problem

Data is being pushed to BigQuery on every run but nobody uses it unless an engineer manually opens Looker Studio. The plugin currently has no way to proactively communicate site health to the person responsible for the site.

This means the tool is invisible between engagements. There is no recurring proof that it is working, and no mechanism to justify a retainer.

## What to build

Every Monday morning, send a one-page summary report to a configured recipient — no engineer involvement required. The report should be readable by a non-technical person.

**Report contents (proposed):**
- DB health status (green / amber / red) with a one-line summary
- Order itemmeta size — this week vs last week (trend)
- Archival signal status — active or clear, with plain-English explanation
- Query execution time vs baseline — improving / stable / degrading
- Autoloaded options size — trend week over week
- Last archival date (if applicable)

**Delivery:**
- Email to a configured address (set in ProPerf → Settings)
- Triggered by WP-Cron on a weekly schedule

## Why this matters for selling the product

A store owner or account manager receives a weekly email proving the plugin is watching their site. They do not need to log into WordPress to know things are fine. When something is wrong, the report shows it before anyone has to ask.

This is what turns the plugin from a monitoring tool an engineer checks occasionally into a retainer deliverable a client sees every week.

## Acceptance criteria

- [ ] Weekly report email is sent every Monday to a configured recipient address
- [ ] Recipient address is configurable under ProPerf → Settings
- [ ] Report includes DB health status, itemmeta size trend, archival signal status, QET vs baseline, and autoload trend
- [ ] Report is readable by a non-technical person — no raw numbers without context
- [ ] If no data exists for the current week, report is skipped (no blank/empty email sent)
- [ ] Report is triggered by WP-Cron — no manual action required

## Related

- #11 — server-side metrics (more data to include in the report once collected)
- #34 / #35 — alert system (report and alerts are separate; report is a scheduled digest, alerts are event-driven)