Guys, since you are not willing to give us SIDE/MID control natively into the devices, what about adding “simply” a helper device functioning just like Voxengo MSED. So you can:
swap (“enocde/decode”) L/R <-> SIDE/MID
gainL/MID or R/SIDE
mute L/MID or R/SIDE
pan LEFT/MID and RIGHT/SIDE individually, post encode
I think this would not be too time consuming. You have all interpolation algorithms and the encoding/decoding is not worth mentioning.
And I also think its not wrong to have a side-enocder-tool.
How much time does this cost you?