bumping this one, i think it’s a cool idea. just having a “patching” fx matrix for incoming audio makes sense, and is pretty much what i use now. but, extending this to channel interactions like ability to morph by spectrum or realtime granular morphing etc. (that is drawn on the grid on a separate “page” for example) would be quite fun. also, simply selecting channels as control channels for modulation, or sequenced patching states… in this type of system, one voice with some feedback would be well enough 
i did check the pd thread but didn’t find so much about the grid itself, and probably osc is the way to do it. maybe grid grouped with another controller makes sense, preferably with some indication of channel levels.