Much of this can be explained because he was at the right place at the right time. Two very important reasons why Linux is so successful are GNU and the GPLv2. Any decent kernel could have taken off at that time. Linux was the first that actually worked.
Now the ability of Linus to retain leadership is remarkable. Just, let's not forget that he got a hell of a head start with the FSF's work. Both technical (GNU) and legal (GPL).
Linux is not an OS. It's a kernel. The OS was GNU, and Linux was only the missing piece.
I maintain that any kernel that worked with GNU would have taken off.
However, what I got wrong: in the first years, Minix was not free. Users had to either buy the book, or pay a license fee of $69.
If Minix (or at least the Minix kernel), used the GPLv2 from the start, I wouldn't bet on Linux.
Also: I was thinking he's remained the leader of this very successful project, for a long time - he deserves credit for that. But perhaps that's a part of software's durable unfair advantage, that e.g. made so much money for Oracle, Microsoft. Similarly, BDFL, as for perl, python, ruby etc tend to stay there. So, perhaps it's more just not stuffing it up, than special credit?
What do ypu think of how esr explained linus's success in the cathedral and the bazaar?
I've worked with a few people I'd consider brilliant. But none whose accomplishments really compare to those of Linus. He's more or less a sample size of one as far as I can see in terms of building out such successful software on such a large community and without being inside a megacorp.
I've worked with many more people who feel the need to be involved, steer projects towards failures or dead ends, who seem to live by the CIA manual on field sabotage and have to be managed to not cause a net detrimental effect. Some of those were really nice people too.
I'm amazed at how he's built Linux without more clashes, but if he feels he can be a nicer guy about some of it and still do the job effectively, more power and credit to him.