It works relatively well on a beefy BBB instance. Jitsi from my experience craps out on low-bandwidth scenarios, but I've attended BBB meetings of respectable size with no issue.
WebRTC as currently implemented has very hard and by its nature has very low audience caps. There's a reason why WebRTC-to-HLS gateways exist; WebRTC by itself isn't substitutable, and if you're going to act like this I'd think there's a pretty significant onus upon you to provide something substitutable.
The substitutable thing is "another content provider that does HLS or equivalent", and we're back where we started, because doing so independently, as discussed in another subthread, either costs a lot of money or comes with compromising tradeoffs, as well as a support burden that somebody has to pay for and do.
Independent web video is mostly awful. The best answers are bad ones. Sorry.
It works relatively well on a beefy BBB instance. Jitsi from my experience craps out on low-bandwidth scenarios, but I've attended BBB meetings of respectable size with no issue.