But thinking about it, it started to make sense to me. Red Hat always organizes upstream communities. The freely available, unbranded (as in, not Red Hat-branded) software is always ahead in terms of features, and possibly lagging in stability: oVirt, Ansible, Foreman, and many others.
So in this way, CentOS was always an exception. Not driving innovation, but taking tremendous resources to keep decade-long promises.
Now I understand that CentOS was not launched but acquired by Red Hat, and in turn how they broke a serious LTS commitment shortly after release. And sure, the trust is damaged not necessarily because they withdrew a free project, but because how they did it.
But for those of us who manage smaller deployments, it seems like CentOS Stream could be a very nice middle ground between new features introduced in Fedora and the stability of RHEL. Back when CentOS 8 was awaiting release, my team decided to install Fedora on our (less than 5) servers, but now eyeing Stream as a viable alternative.
Burn me once shame on you, burn me twice shame on me.
Killing a already made promise is just beyond rational.