We've hit peak audio (best reproduction, no restrictions on usage) and it's only downhill from here.
I look forward to my 2025 speakers that only work for an hour a day unless I pay for extra time credits.
"Do you wish to play a) music b) music and local radio c) music, local radio and podcasts [BEST VALUE]?"
Modern digital technology makes this so much easier and more. There are excellent digital power amps, where you have power amplification on one chip integrated with the DAC. Just looking at some of the features of a simple Homepod, there are exciting technologies involved, which could improve also higher end audio. First is driving the bass speaker in a feedback loop. In classical audio technology, you would have a power signal and have to rely on the speaker to transfer this into motion with as little as possible of distortion. Which required the speaker to behave like a perfect spring for different frequencies. So this is a very difficult task, making the speaker expensive and often imperfect. Homepods drive their bass speakers in a feedback loop, the desired position of the membrane is calculated and an electric circuit drives it into that position. It does no longer depend on the mechanical properties of the membrane, also allowing for much higher motion range than in a classical speaker. Also, the active monitoring of the room acoustics with several directional microphones is an improvement vs. classical set up amplifiers, even if they had a microphone input for set up.
So I don't see us hitting peak audio at all yet, that makes for exciting times for music lovers.
Distortion levels from speakers are still far from negligible-- by negligible I mean so low as to be extremely likely to be inaudible.
High power amplifiers are just now reaching a point where there are efficient amplifiers with negligible distortion; though they've been pretty good for a fairly long time.
There is a lot that can be done in terms of immersive spatial audio, unfortunately the trouble and cost of installing an array of speakers ... limits deployment. :)
There is also a lot that can be done to use DSP to ameliorate poor room acoustics, this stuff exists, but it isn't super widely deployed.
Nah, there are plenty of other things, like positional audio. I remember (through a pleasant haze of nostalgia) my old A3D-based sound card in the 90s as being even better than modern EAX stuff.
Potential improvement in that area can be both through simply adding speakers, and through tuning per-speaker output to forge better audio-cues. (To wit, Head Related Transfer Functions applied to headphones, possibly even with custom parameters for different peoples' heads.)
So, nothing, at the end of the day?
It does plenty for music.