After testing and undoing the sync modification for the 7.11A firmware, one board stopped working (no video output). Comparing some scope signals with a working board the problem was found; a not working 74HCT04 produced a distorted clock for the EF9345. Replacing it with a 74LS04 fixed the issue.

The 74HCT04 was tested with my TTL-tester, and functionally it works OK. Apparantly it doesn't want to work with the 12 MHz crystal! It might have oscillating at 24 MHz, but with other signals interfering.

Below the video signal at the base of the Q1 transistor.