Most modern voice applications give a pretty clear visual indicator about who is making noise. I think in your scenario, the person echoing the audio would be quickly identified and told to mute themselves. If they were the only ones playing your audio back after a delay, people would suspect that they are the ones with the connection issue.

However, this was pretty common for a prank, especially 10-15 years ago on services like Ventrilo which resulted in the then-popular Ventrilo Harassment series. Most of the people who got pranked in those didn't get a visual indicator of who was playing sound. The situation you describe can be seen in this video titled "Ventertainment - Nerd Confusion 3: The Ring":

https://www.youtube.com/watch?v=b5AkIfgioA4

I worked at a radio station and would do this sometimes with a tape deck (Revox) that could record and playback at the same time. Was always fun to trip up the presenters, it's almost impossible to talk that way.

I'd never do it on the air by the way, I'd use an aux bus while taking to them over the TalkBack while music was playing.

This was a hilarious clip. Brought me back to 15 years ago, messing around with people on Ventrilo while playing World of Warcraft. Many, many lost nights and days playing that addictive game. Comic relief thru Ventrilo harassment was necessary...

"I've got baaallls of steel"

https://youtu.be/FvL-WGcoBis

Most meeting software I've seen shows talkers, making it very obvious who's causing the echo.

I've never seen local echo on a video conference nor do I really see a realistic path that could reasonably cause this. Regular echo, on the other hand, is common.

I’d attribute that to the other person playing the audio out of their computer speakers while using a mic that is not directional enough to filter it out.

I generally consider it rude to not use headphones in a meeting. The primary issue is you can't have simultaneous talking when someone isn't using headphones, since most applications auto-cancel audio input when they detect feedback. You then have to switch to turn-taking which isn't always suited for certain conversations.

I dearly wish this could be an actual norm.

> playing back my own audio with a delay

Sounds like this device: https://www.google.com/search?q=ig+nobel+speech+jammer