The idea that QWERTY was designed to slow us down, and that DVORAK is significantly better are both surprisingly long standing urban legends. Liebowitz and Margolis wrote the definitive article debunking this (in 1990!) with loads of research. The jamming issues were sorted out before QWERTY became standard, and it actually won out over a number of other layouts over a period of years. Additionally, the studies show DVORAK is better generally came from Dvorak himself. Independent studies (like one done in 1956) show there's no appreciable difference. This is probably the main reason DVORAK hasn't really made much ground, even though it's been around since 1936.
http://www.utdallas.edu/~liebowit/keys1.html (1990)
They wrote a follow up in 1996 showing how this myth keeps propagating, and how authors keep referring to each other, making the myth sound legitimate. Having 25 citations certainly makes it sound like it's true.
http://reason.com/archives/1996/06/01/typing-errors (1996)
My guess is DVORAK users may have some form of "sunk cost" bias, considering they spent the time and energy converting to the new layout. Possibly some affirmation bias on the old studies. Not sure how else you could justify the costs of using a non-standard keyboard with no conclusively proven gains in speed.