Measured phone UI coverage of native source pixels

Twenty phone views were measured at the same ten natural source/profile sites as the physical Hex-density batch,using the unchanged isolated production renderer for commit d111bb3. All twenty actual foreground transforms match their earlier controls. No source image,source permission,world feature or renderer response was replaced.

The method compares the actual drawn source alpha with a UI-only matte rendered over black and white backgrounds. Only source pixels with alpha at least128 and estimated UI opacity at least0.85 contribute to the reported overlap. This measures strongly opaque UI covering the drawn source body;it is more specific than testing intersection with a DOM bounding rectangle. Temporary matte backgrounds are measurement instrumentation,not an artwork change or a normal artistic-approval frame. No matte PNG or raw RGBA buffer is retained.

Two views show measured overlap:

The other eighteen measured views,including all ten Hex views,have zero overlap at these thresholds. This does not prove full visibility of alpha below128,all possible camera positions,all display sizes,backdrop-filter effects or every final scene pixel. Existing artistic,ground,scale,state,lifecycle and full-performance gates remain. No new asset approval is granted.

The two remaining Region overlaps are a specific framing/navigation issue. A normal camera-pan follow-up at unchanged native width is the next check;these measurements do not justify repainting or shrinking the original city artwork,or shifting every Hex portrait when the measured Hex views are already clear.

Eleven of the twenty complete map hashes match the earlier control screenshots. Nine differ despite identical foreground transforms. Their cause remains unresolved;do not attribute the differences to source artwork or erase the earlier terrain/Hex nondeterminism gate. The original source-only/native checks and physical-density results remain separate evidence.

The server and every browser context closed successfully. The owned disposable world was retired after terminal and path checks. Compact per-asset findings,source/runtime hashes and two decisive Walled Gate phone witnesses remain. Artist files,both Citadel hamlets,production geography and persistent saves are unchanged.

Evidence: per-view results results.json, validation summary validation-summary.json, owned resource closure owned-resources-terminal.json, retirement receipt owned-world-retirement.json. These are local follow-up findings;they are not new production-world observations.