I have no plans to stop working on this project any time soon. Having it work as a ChromeCast Audio replacement was one of my initial thoughts back in 2019, and while the scope has slightly increased, that initial goal hasn't been altered.
After some feedback, I am prioritising the implementing of multi-target streaming (multiple speakers/TVs) so that this will exist (in some form) in the first release.
Another of the goals is for the whole system to be dead-simple to setup and use. Heavens know I have wasted enough of my life trying to make half-broken projects somewhat do what they were supposed to do. I'd rather my own projects are not among those :)
(I'm the NymphCast author)
Certainly as an end user the MPD ecosystem looks to be "stagnant" to put it generously, many of the clients I tried demonstrated curious behavior, and at least for audio there don't seem to be any obvious alternative standards emerging to replace it yet.
Doing this seemed to be unreasonably difficult. The closest thing to a solution I identified was using Mopidy along with Icecast, where Mopidy is the control plane and Icecast is the audio sink, and all client devices simply permanently connected to the shoutcast stream.
This "worked" but it is very frustrating for a simple use case. Mopidy used to be a MPD server implementation, but it has deprecated that in favor of its own API, and the client options were very limited. And nevermind video, this was just audio...
Anyway, I feel like NymphCast is a very welcome development, and I really hope that Maya will continue to work on it.