What you describe is not really surround, but distribution, so I would definately just solve the whole thing with matrix outs, and not the speaker processor 
as it’s possible to eq and delay the matrix channels separately it should be no problem, assuming you are not using surround fx.
And if you need some fx in surround, I would do these with automation or control from the laptop you mentioned, and a 5-6 MADI channels routed to the different matrixes. You should easily be able to program/automate some fx this way.
