The GNOME Calendar maintainers bring up the fact that Linux Mint ships a nearly 2 year old version of GNOME Calendar to its users and still keep the GNOME branding and support links.
The Linux Mint team don't see an issue with shipping old, unmaintained software and would rather ignore the issue and pretend it doesn't exist. The recent releases of Linux Mint have downgraded several pieces of software to their older GNOME 3 versions which re-introduce fixed bugs and removes newer features. This was done to support their theming engine, despite the fact that their theming is already several releases too old for GNOME apps.
In addition, they utilize several old GNOME related system components that apps rely on such as GNOME Calendar so even installing a newer version of Calendar wouldn't be enough.
This ties into a much larger issue of "Long Term Support" desktop operating systems where software can be pulled back for nearly 5 years (or even longer under a contract). Necessary fixes that are too complicated to be patched back in (or count as features) are then missed by users. Users end up using years-old flagship desktops (GNOME and KDE) and get their first impression of those projects from versions that none of the upstream maintainers support.
It's a shitty situation all around: I personally think it's incredibly important to use the latest release of software that's supported by upstream. GNOME does not actively support older versions of its app and shell from more than a release ago and distributions shipping these older GNOME releases can only cause damage to GNOME's reputation.
If this isn't possible, then I don't think it's right to still keep the branding and trademark of the application, much less keep the original support links intact.
Stuff like this is why I've been telling people not to use Mint or Debian or whatever when they've talked about switching to Linux on Hexbear. Fedora has a fine release schedule that doesn't usually break things as long as you wait 2 weeks to upgrade to a new major version every 6 months. Though they are not immune, as the other linked article shows in their back and forth with OBS. Fedora flatpaks were always bad.
People are missing out on improved software and bug fixes, newer mesa drivers that can significantly boost game performance and compatibility, new kernel improvements that increase overall speed, wayland support, etc. You can pin packages on your own or use containers if you're someone who actually needs your software to stay on a certain version. Both Bazzite and CachyOS, two of the more popular gamer options, have LTS kernel boot options by default too. I don't think there's a lot of good reasons the average person who is relatively tech literate should be using LTS distros, even if they don't play games at all.
Someone else mentioned it in relation to Duckstation, but there's been plenty of situations like this before. Someone releases an app that they only distribute as flatpak, appimage, whatever. Someone else makes their own unsupported package for it and doesn't do anything to direct issues away from the main developer. Main developer gets loads of issues and complaints they have to keep shutting down from people who don't understand the difference. Developer gets rightfully annoyed. Same as it ever was.
Fedora flatpaks are such a baffling choice I dont even know why they exist.
Fedora seems technically strong but the association with Red Hat and its governance making iffy decisions seems like a longer term risk. And Red Hat are a big MIC contractor.
Opensuse tumbleweed is basically the only alternative to fedora which I used to be happy running but zypper is a single threaded slow as fuck pile of poo which makes updating the OS a chore. If they sorted that it would be a strong fedora contender with a cuter logo
There's also KDE Linux, PopOS (though Cosmic is basically still beta software), etc. Just anything that actually uses updated packages. They all have pros and cons but I think the cons of most of them are significantly less than the situation described by this post.