Powershell uses a structured data pipeline, as does elvish and nushell and many others. I don't think fish does, it's trying to be mostly POSIX-compliant except where they don't agree with POSIX. I might be wrong about that it's been a long time since I used fish.

Feel free to keep reading my ramblings but the landing pages for both elvish and nushell that I linked above illustrate the power of structured data pipelines, in particular nushell:

Accomplishing something like this with a flat text pipeline like bash would be much longer and far less readable.

POSIX-compliant shells use plain text in their pipeline, which means that any output you pipe from one command to the next has to be done so in either plain text or a format the consuming command can process. What that means in practice is that unless the consuming command has been designed specifically to consume the output from the preceding command in the pipe, users have to do a lot of string manipulation with tools like awk, sed, or grep to structure the output from one command into a format that can be consumed by the next.

Where it really shines is when interacting with APIs that expect to use structured data in their interfaces, as you can just pipe data from one to the next without worrying about restructuring it. It's why PowerShell was created--unlike POSIX-compliant OSes, Windows is built on the component object model, and every API built into the OS already used structured data.

The web is built on structured data (json, xml) as well, and that lucky coincidence made powershell a much more useful shell for over-the-wire API interaction than bash et al.