I’ve just abandoned the aggregate option.
I’m currently running 7 outs into the modular (3 x audio, 4 x CV) then bringing 4 feeds from that back into Max through the ES8s 4 ins. I then mix / mangle that more in Max and send a mono out (lowered in level) through the remaining ES8 output for monitoring.
I’m recording everything using multichannel files in sfrecord~ for chopping up and mixing later - all working great.
Feels like there’s loads of scope in that setup, and it feels like a workflow of sorts…