Welp, sorry to say this, but it looks like I have failed (again).
I was literally getting ready to record a new video as well ;_;
I think she may have blown a fuse or fried a chip in one of her shoulder motors.
Suddenly the behaviour of the connected benchtop PSU changed completely. I couldn't increase voltage without first having at least 7-8 Amps. And a small increase in voltage now causes a large increase in amperage - which never happened before. Before this, the PSU would only pull about 4A even at maximum recommended power (24V).
This fault occurred while testing new movement sequences (I forgot to lower the acceleration of her shoulder motors, so I think they took too much power). But I lack the courage/knowledge to go poking inside my PSU since this is one way you can actually get some nasty electric shocks/potentially die.
Plus, at one point her right arm completely spazzed out, almost K-Oed me and pulled a bunch of wires (which were wrapped and will take some time to re-attach).
I think one needs a lot more knowledge of electricals than I have to pull this kind of project off. And ideally practical help from at least one other person who knows what he/she is doing.
The servo sequencers are certainly the way to go. It's just that before this, I'd never managed to program Sophie to move very much, so her power supply and motors were never really placed under any strain.
Solve one problem, and another even worse one crops up a few hours later.