It doesn't sound crazy, because he spends 90% of the essay more or less arguing semantics. As he says, AIDS isn't a disease. It's a syndrome. And the diseases which do the damage aren't caused by the HIV virus. He just points that out like it's something which no-one else is brave enough to mention, when he's really just attacking strawmen.

The really crazy bits are when he talks about whether or not HIV can damage the immune system, by killing T-Cells. I counted 8 mentions of the word "immune", and 7 mentions of the word "T-cell"; and he only mentions them halfway through the document.

Because of the complicated chain of causality (untreated HIV eventually wipes out the immune system, which causes a whole bunch of weird diseases to take hold), it's easy to write thousands of words about why rare diseases (which are called AIDS, in HIV positive patients) are only correlated with HIV, not necessarily caused by it. But in the middle of the article, the crazy bits are there.

It's like reading a long rambling article about why CO2 doesn't necessarily cause global warming (which has a grain of truth to it - it's a little more complicated than that with the role of positive and negative feedback effects), and halfway through the article the author suddenly mentions the greenhouse effect for the first time, then says it's unfounded (which changes the article from slightly crackpot to stark raving mad) .