I read a fair amount of news during the day. It's mostly fed to me through XML feeds that I subscribe to with Google Reader. Very often, the sites to which I subscribe link to news from other sites; every one in a while, the New York Times' website will be the destination of a link in the article I'm reading.
The most recent article I read is one on Asperger's Syndrome, talking about how a competitor in America's Next Top Model has it.
Now, the article wasn't that interesting to me, so I skimmed it. When I skim on the computer, I help myself focus on the text by highlighting the paragraph that I'm reading. It's the digital equivalent to following one's finger while reading a book, or using a piece of paper to keep track of where one is reading. The easiest way to do this is just double click the block of text you want to read — go ahead, try it on the next paragraph.
I have built up a habitual double-click computer reading style. I can't help but do it, and I'm convinced that it gives me more comprehension in less time. Unfortunately, the New York Times' site has an awful, ridiculous feature tied to double-clicks: it opens up a NYTimes.com search (in a new window) with the words that you clicked as the query.
I'd wager that 95% of their readership doesn't ever intentionally double-click the article text. Of those who do, most probably expect the standard browser behavior: highlight this text, allowing me to copy it easily. For the really, really small group who (like me) double click every few seconds, it's nigh on impossible to read an article. This sounds like an idea that some developer though would be really cool and oh-so-user-friendly — instead, it's invisible at best, and a real hindrance at worst.