Indeed! I rather like the idea that solitons are something like the simplest self-propagating 'things' in any medium/computation-reigeme, so gliders in Game-of-Life might qualify and in more complex/subtle systems they can have more complex behaviours as well (like bacteria, or flies? Hehe) Here's a fun example I made in gollygang/Ready (and Houdini) of PDE solitons that spin around with rippling wakes:

https://youtu.be/edNG2EJWQwQ?si=8b3YsHCDIaj0C7Pa

Would sound waves then count as solitons?

I believe so, although the way I usually think about Solitons is like a single packet.. so just one cycle of a wave. Continuous sound could probably be thought of as a continuous stream of solitons (I think ppl call them phonons when it's sound though). I haven't studied PDEs nor solitons in a formal way I just love playing with them. Gray Scott with History and Wave (a formula I contributed to Gollygang/Ready) supports many fascinating soliton behaviours. Here's 25mins of one of the strangest parameter settings I've found:

https://youtu.be/Naj_J8aznyk?si=Da3A3iTz8rN9qrgq

Only if they retain their original shape. The point is not that any wave is a soliton, but a soliton never changes shape as it moves (through time, a medium, or whatever). The soliton can decrease in amplitude, and expand in width, but otherwise remains the same.

A pure, single Gaussian hump is the soliton for homogenous linear media. If you create an audible with the spectral shape of a Gaussian (and therefore also the time shape), it might get quieter as it moves across the room, and longer, but will still "sound" the same.

Wow! I'd never heard of gollygang/ready before, but it looks extremely cool. Thanks for sharing.

No probs, I've had a great deal of fun playing with it indeed! Hope you do too!