i just ran into 2 very interesting articles by Sacha. They are:
the first one is fun. Go there, and visit the typing speed site she linked, and test your own speed.
she types 110 wpm, amazing. I only got 80 wpm. Also, she types Dvorak, yay! (since 2003) But she also retained her QWERTY, at 88 wpm. wow! My QWERTY is about 30 wpm.
i've seen a few typing speed sites over the years, their quality varies significantly. The one she found is the fantastic, probably the best.
also, note that different sites give different results of words-per-minute, because different input has significant impact on your speed. You are much faster when typing chat-like text, but slower when typing journalism essay, and much more slower if you were to type Shakespeare. Also, typing test's duration effects your score. You might be 70 wpm for 1 min, but 50 wpm for 10 min.
this means, if you hear programer claim 60 wpm, you need to know what's the test exactly. The reported wpm difference sometimes differ as much as ±30 wpm. (i know i'm 80 〜 90 wpm thru-out my life, but i've came across a site that rated me 120 wpm.)
her second article is a little research on different statistics on typing speed. It answers questions such as: can you type “asdf asdf asdf asdf” faster or “dfjh dfjh dfjh dfjh”? (that's “rolling fingers” vs alternating hands) Or, which is faster: “jjjj” or “kkkk”? (repeating pointing finger vs repeating index finger.) It includes a little snippet of emacs lisp, so you can test yourself.
why's that interesting? because it's a science based experiment, for issues in keyboard layout. A actual test that you can run for yourself. I haven't seen such study elsewhere. It's commonly said that Dvorak favors alternating hands and Colemak favor rolling motion.
Here's 2 excellent typing games, one of them is mentioned by sachachua.