Extrapolate the chord progressions into well voice-led melodies rather than to arpeggios.
251 triad example:
D F A
G B D
C E G
becomes:
F A D
G B D
G C E
(To minimize the distance a voice travels so it reads as melodic rather than stochastic)
Now derive the sequence by reading vertically rather than horizontally:
Seq 1: F G G
Seq 2: A B C
Seq 3: D D E
(for contrast, I understood that your post, given the 2-5-1 in C Major input, would yield the sequences:
D F A
G B D
C E G)
I’m anxious about 1: relaying something that was mind-blowing to me but may be obvious to you and 2: relaying the information in a way that does not emphasize the important parts of the approach (chords are stacks of melodies; melodies are not simply vehicles for chords AND use simple sequences; they’re more primed for live modulation).
I’m also sure you can see that, if you’re building your chords in more elaborate ways // with this application as an end goal, there’s a lot of freedom // it’s a lot easier to change the sequences to taste without destroying the whole piece (whereas changing arps can be tricky since they’re so… Crystalline).
I thought arps were the way to approach chords in modular, too, but then when I started thinking of it as an ensemble rather than a glitter box, it made more sense to tackle melody directly (even though it was scary at first).
[also this was a quick and simple example with no extensions I could do quickly – I don’t mean to suggest that I’ve found a great way to voice lead this particular progression]