The article talks about why Anubis shouldn't work but not much about why Anubis has worked in many cases. Truly a mystery.

As finitebanjo said, the article doesn't say why Anubis works.

Those crawlers, Anubis is made against, are the sort of trash that'll DDoS your site due to the sheer amount of requests. And while the calculations are cheap to perform once, if you're requesting access over and over, their cost piles up.

Is this perfect? No. Better than letting some "A""I" company to DDoS you? Yes.

I'll go further - if I had a site I'd be deploying it alongside Nepenthes. Fuck this shit.