I'm sorry, but I have to agree with Eliezer; I've known the details behind a half dozen stories or so personally, and in almost every case in the MSM there have been factual errors reported as truth.
One time is an anecdote; two times starts to be compelling. 6/6 times has a reasonable level of statistical significance.
FWIW, I'm somewhat convinced that a majority of computer programmers would fall under the general umbrella of "bumbling," though I don't think I'd use that term. "Marginally competent" might fit the bill better.