Right, update time!
Between my day job and having a small kid at home, I’ve only been able to spend scattered moments on this (hence the two-month silence), but I have tried to playtest the prototype as much as possible, and: I reckon the concept works, for what it is. I’m also happy with the pan law I chose (3 dB). The positions definitely need to be changed, though. I’ve experimented by manually tweaking per-channel gain of a synthesized stereo sweep (in Audacity), and have something I think will work better, and that I will use for the second revision, along with a few other minor tweaks.
Playtesting it has also given me a chance to think about additional features. Here’s the thing: I want this to perform a single, specific task – mixing seven channels down to two – and to be good at that, without locking it into a specific use. I want to it be small and inexpensive enough that you could use a few of them throughout a patch without thinking too much about it. Things like attenuators, switches, a line-level stereo output jack, or even expander pins, would all be useful in some circumstances, but they all add size and cost (and complexity). I wasn’t sure that a fixed-position panning mixer could be useful without any of those features, but now I am.
Which is to say, I don’t expect to make any radical changes to the design, except possibly moving the output jacks to the opposite ends of the panel, as suggested by @laborcamp, since that definitely strengthens the metaphor behind the interface – if I can find a way to do it without making a mess of the PCB layout. I hope to get time to finish the second revision this week, and will then order new boards to build and verify. I have a few sketches for panel designs, but nothing final yet.
Oh, and I also settled on a name. A minimal mixer needs a minimal name; meet the 7:2.