[lnanek2]

It's stupid and ugly, but I've seen studies before where click here simply got more clicks. It's simply more effective to tell people straight out to click something, than to hope they pay enough attention to an underlined noun and have enough interest in it to investigate it without any prompting, etc..

This blogger has mentioned before that he sees no reason to A/B test because he thinks an experienced designer knows better anyway. So I can totally see him recommending something that performs worse in an A/B test like this.