[RUM] LCP performance regression: '/' view degraded +61.7% in last 24h (4320ms p75)

## Summary

Datadog RUM detected a significant Largest Contentful Paint (LCP) regression on the Studio landing view (`/`) over the last 24 hours (2026-07-01 14:12 → 2026-07-02 14:12 UTC).

## Metrics

| Metric | Current p75 | Previous p75 | Change |
|--------|-------------|--------------|--------|
| LCP | **4320ms** | 2671ms | +61.7% ⚠️ |
| FCP | 2978ms | 2434ms | +22.3% |

LCP is now above Google's 4000ms "poor" threshold.

## Anomaly Windows (selected)

| Window | Avg LCP | Peak LCP |
|--------|---------|---------|
| 14:50–15:00 UTC | 10,043ms | 14,703ms |
| 17:35–18:20 UTC | 5,133ms | 5,451ms |
| 03:00–04:20 UTC | 5,003ms | 5,760ms |
| 08:55–10:40 UTC | 6,714ms | 8,372ms |

FCP tracked closely with LCP throughout all windows, suggesting the issue is with initial resource loading rather than JS execution time.

## Recommended Investigation

1. Run `get_rum_insight` with `insight_name=aggregated_waterfall` on the `/` view to identify which resources are causing LCP delay.
2. Check for large asset changes (images, fonts, scripts) deployed around 2026-07-01 14:00 UTC.
3. Compare resource sizes and CDN cache behavior day-over-day.

## Source

Datadog RUM — Studio app, daily automated review 2026-07-02.

Start by running get_rum_insight with insight_name=aggregated_waterfall for the / view. Then compare images, fonts, scripts, resource sizes, and CDN cache behavior with the deployment around 2026-07-01 14:00 UTC. Done means identifying the resources responsible for the LCP delay and correlating them with the regression.

Written by the indexing model from the issue text.