I agree that touch-typing and greater typing speed is beneficial... hard to argue it. But the typing test definitely favors the touch-typist, as you are copying random text. The touch-typist never has to take his eyes off the screen... big, big benefit in this test. The non-typist has to find his place in this large text every time he brings his eyes back up... and it takes longer to to absorb the next piece of text he is going to copy... you have to remember capitalization, punctuation, and spelling of a text that is not your own thoughts.
In coding, however, you are typing what is in your head already... you are not having to go back and forth with your vision, the text you are writing is coming from in your head, and you are not concerned about replicating something on your screen. I'm not making the case that touch-typing is wrong (that would be dumb), only that these standard tests are not a true measure of programming typing speed... I think the hunt-and-peck crowd would at least fair a little better in that circumstance.
However... this has pushed me top go back and re-learn touch-typing... it would be a benefit. And if there are any other developers or managers that are judging my performance as a programmer on my lack of this skill (whether warranted or not), it is worth taking the time to learn.