I read a couple of comments on another blog that put me in mind of this story.
This is about a coworker I had. I've written about him before. He's the first person I had to fire (sadly, not the last). He was/is super-smart, but that didn't translate into his being able to do the job.
We butted heads on a few things. Before I was his supervisor, this happened, which wasn't him and me butting heads, but anyway: The EPA requires certain testing to be done on soil and water around underground storage tanks containing gasoline and diesel. They started requiring the samples from around gasoline tanks to be tested for benzene, ethylbenzene, toluene, and xylene by purge-and-trap, GC/PID (PID is photo-ionization detector) and gasoline-range organics by purge-and-trap, GC/FID (FID is flame-ionization detector.) (GC is gas chromatograph.) PID is sensitive to conjugation and it does well with aromatic chemicals like benzene, etc., while FID gets all of the hydrocarbons. EPA had started adding MTBE to the list of chemicals to be tested for and the engineers asked that we add it to the PID analysis, as other labs were doing. The coworker told the engineers that that would not work. There's a number associated with molecules - can't think what it could have been - but he could look it up on a table, and did; and typically for molecules whose number was less than X, PID didn't work. The engineers said that other labs were analyzing MTBE by GC/PID. The coworker said it simply wasn't possible. "We have to do it." "It won't work." This went back and forth and they were in a complete standoff when I walked in upon this. I wasn't the guy's boss yet, but I said to him, "Why don't you just go and make a standard with MTBE, and run it; and when you don't get a peak on the PID, we won't have to talk about it anymore."
This was acceptable to him. He went and made a standard with MTBE and ran it. Well, you know what happened, ha ha. He added MTBE to his BTEX standard and we heard no more about that table with the numbers.
I am more pragmatic than theoretic. There's a time and a place for theory, but empirical evidence trumps theoretical every time. Trust but verify, I always say. You have less of a chance of looking like an idiot later.
But the coworker had problems submitting his internal set of requirements to the requirements of the job; and most of these were bona-fide necessities. He did things his way, and if they were the right way, fine, and if they weren't, the rest of us needed to get over it. Eventually he had to go because the liability was too great. I was very irritated to learn that the lab manager who retired before I was promoted, when he heard about my promotion, immediately asked whoever told him about it if that person still had his job. He knew the guy needed to go and wouldn't take care of it himself. I guess he liked him. I liked him too. Liking him wasn't the issue, it was the responsibility I had to all the other people who were doing their jobs correctly, including the tedious parts, and who were vulnerable to losing those jobs if we lost work due to this guy.
This coworker told me a story once that I thought later was very significant. When he was a third-grader his mother gave him an algebra textbook. Algebra fascinated the hell out of him, and so he blew off his third-grade arithmetic, never did his assignments and so forth, and received an "F" for the course. They were going to make him repeat the grade until his mother took him in with his algebra textbook and demonstrated that he was way ahead of 3rd grade, and then they passed him on to the next grade and left him alone about the math.
I tend to think that this was handled incorrectly, because what the boy apparently internalized was that rules and structure were for people of normal intelligence. If you're smart enough, you don't have to do the tedious dumb stuff. Except that you do; everybody does. I think his mother probably should either have tried to get his 3rd grade math assignments changed to match his capabilities, or else have explained to him that you just can't soar with the eagles 24/7/365. As smart as he was, he had a real disability when it came to functioning in the workplace.