Ah, the great mystery. If you find out, let me know.
Technically speaking, there are legitimate disadvantages to Linux compared to Android, iOS, and macOS (and Windows to a lesser extent), some of which I’ve previously brought up on Fedora’s discussion forum:
However, it’s important to understand the context of these supposed deficiencies within the larger threat landscape. For example, the average person’s greatest risks by far are phishing attacks, weak passwords, and drive-by malware downloads. Your OS won’t help you with the first two, and in the case of malware downloads, simply not using Windows is probably the greatest thing you could do to reduce your attack surface.
The “problem” with Linux is basically that it would probably be easy to develop malware for it. The reality is that despite this, nobody is actually developing malware for Linux, and you’re exceptionally unlikely to ever encounter a Linux exploit in the wild unless somebody sends it directly to you.
If I were a high-risk target—like a journalist reporting in a hostile foreign country, or an executive at a prominent company, for example—I might not be comfortable trusting Linux as my daily-driver operating system. But at that point frankly I would also need to consider many threats which are a bit out of scope for a project like Privacy Guides.
There’s a general rule I like to follow for regular computer users: If an exploit is not being used in the wild, you do not need to worry about it.
There is value in posts like the one you linked which describe security design flaws, but that value is for developers and cybersecurity professionals, not for people who just use Linux. As someone with an interest in cybersecurity, there are a lot of things I’d like to see Linux improve from a security perspective, but there’s also no reason to also not generally recommend Linux to people in the meantime. The article is an interesting piece for Linux developers to ponder, but frankly I’m tired of people citing it as the reason nobody should use Linux themselves, because it is so irrelevant to the average person.
All of this being said, the criticism of stable release models in that document is actually applicable to regular Linux users, and updating frequently is good advice.