Speaker<-- Master Channel <------------ Group Wav
Group (MCG) \ /
Group A <-------- Group Mic
1. Group Wav to MCG and to Group A
2. Group Mic to Group A
3. Group A to MCG
You should be able to set most of that up with simply creating ChannelGroups and using addGroup. The second output from Group Wav to Group A will require manually adding a connection between the head DSP of Group Wav and the tail DSP of Group A.
Have you seen the dsp_effect_per_speaker example we ship? It demonstrates how to do these kind of connections.