We did not leave a bad asset in a 'we will fix it later' state — we reverted it

Situation

With visual assets, it is tempting to hide a small defect with CSS or another overlay. On the USA board, the new background did not reliably contain the structure the renderer expected. Leaving it in place would have made every later positioning bug ambiguous: was the geometry wrong, or was the image wrong?

Approach

I reverted the faulty asset commit quickly instead of adding workarounds around it. We then split the artwork into layers and used the same layered renderer across countries. A dedicated test also ensured that USA artwork could not render twice, because that failure surfaced immediately during the first integration.

Outcome

Board artwork became part of a structured shared system rather than a special USA exception. The revert looked like a step backwards in the short term, but it prevented more code from being built on a broken assumption.