You'll find out why it's hard in a way that no amount of explaining will do, and you'll find reality less amenable to argumentation than comment boxes.
(If you do try it, make sure you've got two different devices, ideally two different types, and be sure to be streaming from somewhere at least 50ms away on the network, ideally with at least a few ms in variance between requests. Though odds are pretty good that even trying to stream the same song to two identical devices over a local Wifi network would only work most of the time, you'd find...)
https://www.youtube.com/watch?v=CAnEFOr1lP4
hmmm, so i guess it can be done!??!
It just isn't as easy as it looks.
In fact, things that are a great deal more complicated can also be done, such as really fantastic real-time echo cancellation or feedback prevention in a large room with many speakers and many microphones. It just isn't as easy as you would naively think; everything that just leaps to mind (especially if you aren't particularly knowledgeable about signal processing) don't work. You'll pay big bucks for the system that can do that, and there's a reason for that.