It's notable that often we have protocols to tie things together. UPnP/DLNA/Miracast have existed for a long time now & have included standards for media playing, light control, others. And many TV's have even implemented!!
I feel like what has always been trickier is control and coordination. There haven't been many people to step up and own the higher level, the thing that coordinates devices.
It's notable to me that recent successful cast protocols (like Netflix+YouTube's original DIAL, Chromecast) have really simplified. Are bottom up flows. They have allowed the phone (second screen) apps to have a direct way to cast, inside the app. To one other screen, to such an extent that multi-room audio works by first making a group that then presents in the app as one device. (Sonos' proprietary casting is the notable exception.) Where-as historically casting was possible but often used dedicated apps, a UPnP Control Point, that talked to a different UPnP Media Renderer to point it at various UPnP Media Servers. A tri-part system, with control as the third point.
Netflix hypothetically could have used the UPnP to do casting. Phone acts as a Media Server (but serves up one-time-use URLs on the net for actual media) to the movie on question. Then act as a UPnP control point to control media playback. But they made DIAL, which reduced it to a two system network. And to have more resident native apps running (with rest channels open between the devices).
It's interesting seeing what folks do with Google Home and Home Assistant. Orchestrating routines and triggers and what not is kind of neat, in some ways wildly hard to imagine from "back then", but also feels so short in other ways. Is such a primitive level of control. It feels like the task of making actual systems coordinateable and orchestrateable, the real dream of ubicomp, is indeed quite far away, unowned, even though in many cases we have advanced computer protocols afforded to us.