Hello people, first post here.
After reading this post, there's one thing I find incredibly odd about the sequence: time intervals for each message. I know you, Rafal, counted frames and obtained approximated times.
I can't imagine the guys who originally made it thinking "yeah, we have the random segments, and now time intervals will be 0.77, 0.52, 0.33, 0.6 and so on". With that I'm not saying your sequence is wrong.
For what I've read, seems the blinking RangeFinder sequence was 0.5s per LED. And if it wasn't for the 0.33 interval on the "all off" state, I'd assume the display cycles after each 0.5 seconds and as it was an old microcontroller there would be sequences when it would work a bit slower and hence time variations. But that 0.33 at the beginning really bothers me, so I'll assume then they chose decimal values (0.3, 0.5, 0.6 and 0.7) for each message and then the centesimal variations is due to processor slowness.
Easiest and simplest options is usually the right one. For example I wondered if (maybe) they directly wired some outputs to certain segments, instead of multiplexing. But after researching, answer is no. And makes sense, as it would be far more complicated.