He's absolutely certain that RH is masterminding a huge conspiracy to corrupt Fedora from within. This seems like pointless conspiracizing to me, given that ordinary cognitive bias might well do the same, but hey.
(He's by far the least odious of the commenters on LWN. One in particular moved quite rapidly from describing interesting connections between IIRC Mandriva and various Russian investors to frothing against the evil Nazis who took over Ukraine and how they're being fought by the valiant forces of the Eastern Ukrainians and their entirely voluntary Russian forces assisting them with their freely acquired military-grade weapons. That rapidly veered into anti-Semitism (I didn't think it especially likely that Jews would join up with Nazis, but that is apparently what Putin's propaganda machine is now claiming) and then this commenter managed the singular honour of being only the second or third person to ever be permanently banned from commenting on LWN. Classy.
LWN used to be a charming collegiate place to comment, reminiscent of the TeX newsgroups, but not for some time. I think it all broke when the security guys with their contemptuous attitude and violently conspiratorial and adversarial frame of mind turned up. Did you know that Linus is masterminding a conspiracy to weaken the security of Linux because, uh, because, uh, they don't seem to have a plausible motive, but Brad Spengler seems sure that he is. His reasoning appears to be that Linus doesn't instantly do everything Brad and his anonymous PaXTeam cohorts want him to the instant Brad orders it (in the most unpleasant and venomous fashion imaginable): thus, since Brad is for security, Linus must be against it. Also Linus very occasionally removes things from commit messages that would make it too completely obvious that a fix is closing a security hole, to stop people looking through just-pushed commit logs from exploiting holes before a fix comes out in the -stable tree. This is apparently also evil in Brad's book, though it is quite unclear why.
Often, doing what Brad orders would destroy any actual development progress but be really secure -- a typical failure of security guys, they assume that security is absolutely all that ever matters and everything else must be subordinate to it. Other times it is simply impossible -- e.g. at times these people have requested every commit that ever introduces anything of security relevance to be annotated with a CVE number and full exploit, even if it's not known to be security-relevant at the time it's committed: if it's not known, presumably we must either get a time machine or hold off committing it until we know for sure! Since Linus doesn't go round editing other people's commits on an industrial scale and, y'know, has other things to do, that's never going to happen. The security guys would be welcome to do such things, but that would require them not to be arseholes who care so much about being first and so little about actual security of real systems that they have posted hashes of descriptions of vulnerabilities rather than the descriptions, purely so they can prove they were first when someone finds them later. WTF WTF.