Yeah okay, I shouldn't have pushed my luck with two builds. First one was A-Ok, and then I build the next one, running into all sorts of weirdness (but hopefully from a common source).
So I assembelled it and plugged it in, went into HW calibration mode and ran through all the switches, worked fine, except I noticed that I did not get a response from Pot #7, restarted it and confirmed it when it was in running mode. The pot does send CV if I check the pins with a multimeter.
So I looked through the soldering, all looked well, plugged it in again, just to make sure I didn't make some dumb mistake. Now all modular hell broke loose, step-encoder doesn't work and all over the behavior is strange, and in HW-test-mode I get some weirdness.
Picture is worth a 1000 words so instead of writing a million of them here is 40 seconds of video, I'm booting it in HW-test-mode, and trying out the switches for step one. Switches seems to work (also does in running mode), but I get a lot of "noise" as you can see, so somewhere I've made a mistake and I hope this is tied to the fact the encoder isn't working properly.
Things I've checked:
Proper 5v from both voltage regulators (smd one reads 5.06v, 5.00v from transistor-lookalike). All ICs are aligned properly.
With a grain of salt I'd say the resistors and diodes are correct, I've rechecked them, but there's still a possibility I missed something.
All solder joints are soldered and they look alright (by my standards)..
I didn't linger with the iron on the switches, in and out quickly.
Any direction is much appriciated!