No copper changed in this PR. It closes open questions by checking the FFC land pattern and EC11 tab slots against the makers' drawings, and finds the encoder shaft 0.25 mm off the footprint origin. A note said power budgets stayed at 0.5 A when power_trunk.py already carried real 5 A budgets, and the README's +5V figure was out of date. U19's placement rotation was closed because the footprint only fits one way, but with no EasyEDA model, JLC's preview may still draw it wrong. The board, the notes, the vendor drawings and the fab preview were four sources that disagreed.