I had loads of ideas for Pi Musicbox and pretty much none of them saw it past PoC stage. The number one reason for that was the insane amount of extra work and testing required to get things beyond the “it works for me” state. Everyone had different hardware and slightly different use-cases, it’s a huge amount of work integrating multiple things robustly. I found any area that I didn’t personally use myself got left behind very quickly, and that became more and more of the project. Until I was supporting something I wasn’t actually using, and then for me it was game over.
Part of me still thinks there is space for something like this or like Pi Musicbox but I’m so out of touch with what other integrated systems exist these days. I think something built around snapcast (or similar) could be very powerful but again, a huge amount of work for one person.