While I understand the logic of initially thinking “I have multiple OSMC/Kodi/etc., it would be nice if they could play in synchronicity”, I do not think it is a good approach to OSMC multiroom, especially for music.
First, it is a waste of computing resources. The music file only has to be decoded once, then it can be broadcasted as PCM to every room you want to play it. Why decoding it in each room using separate instances of OSMC?
Second, it is a waste of hardware. While a computer like a Pi is fairly cheap, it still means you need a full computer in each room and if it’s headless, you will have to bring a screen and a keyboard every time you want to change the configuration, which is a pain. Plus a Pi is still more expensive than a simple Bluetooth receiver with AptX or an equivalent over Wifi if you want lossless music.
Last, but not least, trying to use multiple instance of OSMC to playback the same music with a latency low enough that it is imperceptible is opening a Pandora box. First, you have to calibrate the network latency to make sure that each time you send a synchronized playback command, each instance start exactly at the same time independently of the network latency, and you have to repeat the calibration each time you start a new song because the latency of a network is not constant. Then you have to calibrate the clock speed of each OSMC instances to make sure that a song that start synchronized doesn’t become desynchronized after 5 minutes (a synchronization message at regular interval is even more complicated because you have to recalibrate the network latency before every synchronization message). Then, since none of the OS used by OSMC are real-time OS, you have to take into account that a process on one the instance may delay the playback, so each instance will have to monitor its playback clock, detect the delay, signal the other instances which all have to play catch up for a while until they eventually resynchronize.
I am not saying it is not feasible, one of my client have just done that with Android TVs but, to give you an idea of the complexity, a team of developers had to rework the Android OS for 3 years before they could make it work reliably. And that’s just one OS. The OSMC and Kodi Teams would probably have to develop custom versions of Android, Linux, Windows and MacOS to make it work, and redo it again each time one of those come up with a new OS.
So why opening a Pandora box when, in fact, OSMC multiroom for music already exists, is cheap and more practical than using separate instances of OSMC: Simply use AptX receivers or Wifi receivers if you prefer lossless music. Basically, all you need is to make sure your Kodi computer has bluetooth, plug the Bluetooth or wifi receivers in the sound systems or amplified speaker of each room, including the room where you have OSMC and pair them with OSMC as the master so it can be paired with multiple receivers. When you want to listen in an additional room, just switch on the sound system in this room, make sure the source where the Bluetooth or wifi receiver is plugged in is selected and they’ll start playing whatever OSMC is playing. It’s that simple. And if you are not in the room where the Kodi computer is, just use the Kore or Yatse remote.
Now, movies are another story. However, I don’t think there is many people who really want to listen to the same movie in different rooms at the same time. For movies, what would be useful would be to pause a movie in one room and resume watching it in another room. And that’s much simpler to implement since it doesn’t require any form of synchronisation; only one instance of OSMC is playing at the same time. And for that part, I don’t understand why it was not already implemented when OSMC and Kodi were called XBMC, considering how easy it is (as long as you accept that one instance of OSMC is the master and will be on 24/7, but a Pi doesn’t use that much electricity when OSMC is idling).