Assume, for the sake of argument, that prior infection reduces your risk of future infection by 95%. That's a far sight better than most the vaccines, enough so that a 2nd infection would be noteworthy on a personal level. But then, since this is an Internet forum, we need to throw the law of large numbers at the situation. Those figures still mean that there are perhaps tens or hundreds of thousands of people out there who are known to have caught it twice. Which then translates into a certainty that there will be quite a few people mentioning it as an interesting anecdote on public forums like this.

Then you throw a novelty effect on top of that. Say that prior infection actually confers 99% immunity. Fewer people will get re-infected, but each of those re-infections is more noteworthy, which makes it more likely that someone will want to talk about it. So the actual rate of re-infection and spread by people who've caught it gets a lot smaller, but the rate of people talking about it on the Internet, and the resulting perception people might infer of how likely it is, might stay about the same.

(Disclaimer: I do mean, "for the sake of argument." Until some firmer science on the subject is forthcoming, AFAIC it's still within the realm of possibility that prior infection doesn't confer much practical immunity at all. This study is promising, yes, but I've been around the block too many times to get all that excited over something that doesn't amount to a whole lot more than an in vitro study.)