And yet, I know a few people who have cursed after updating Ubuntu (things that used to work stopped working). And even switched back to Debian because it doesn't happen as often with Debian.
Feels exactly like Fedora and RHEL/CentOS to me. Either you can test with sufficient people (those people have a higher tendency of breaking) or you cannot evolve and expect others to do exactly that for you.
At some point you have to make the switch, and you don't know if the amount of testing was sufficient or valid (because it simply works for everyone who tested it). You can't measure where it breaks until it does. You seem to be implying that Fedora ships broken code on purpose, but that's an ignorant take on the problem. It was shipped because there was no longer anyone complaining, because it worked for everyone who tested it.
You could say that for the people it broke, they didn't help testing it prior to release. But if you tell that to someone who now has a broken system, good luck making that argument ;-)