You can do it but uh its a bit of a ballache. I played around with using dvco machines as a polysynth, i ended up using an external midi controller (ec4 for the knobs) and ableton with my midi kb for managing the voice allocation.
For example if i tweak env amount on voice 1 i need to go and tweak it on voice 2 and 3 as well. I tried using sounds to save the machine params but a saved sound is independent from the per voice machine parameter settings (stored in the pattern kit) unless you reload the sound on each voice.
It did work but it felt overly elaborate and was a bit of pain to get setup and and work with.
Just doing straight sequencing would be easier (cause u dont have to worry about the machine params per voice) but I feel like using the rytm for polyphonic sequencing - theres just going to be so many little gotchas you really have to look at if its worth the hassle vs just using something designed for poly sequencing. If you had like 3 voices sending midi on the same channel, results could vary synth to synth, could be fine though.
Maybe something like the toiraz squid which seems like it would do everything you want (unsure if it has direct jump but id assume so). I’m sure theres an ios app that would do what you want as well.
This was using dvco as a synth for chord stabs, i mapped the dvco params to performance pads for the 3 voices then used cc’s from the ipad to control the perfs