Well, arguably being the fastest at writing code doesn't mean any of it is actually good.
I once worked with a guy who could crank out massive quantities of code. It made for some pretty amazing proof-of-concept and demo stuff -- but it was utterly unusable in the real world because it was, overall, really badly written code.
It didn't have any robustness. It made stupid architectural decisions. It made unfounded and unsupportable assumptions. It had an awful lot of hardcoded magic because he only solved one use case.
It impressed the heck out of the VPs and the customers who said "oooh, gotta get me some of that". But in trying to turn any of it into production level code, by the time you started trying to fix the glaring holes, remove the things which just simply couldn't work in the real world, and take out some of the shortcuts and voodoo -- there was very often nothing workable left.
This was coupled with the fact that he couldn't/wouldn't debug his own code, couldn't/wouldn't go in and make even simple modifications to it without breaking it, and had quite likely moved onto something else new and shiny and couldn't be persuaded to actually fix or maintain previous projects.
He left a wake of crap code and demos, but almost nothing which could be used in the real world. Which means everyone around him started saying "look, if you want to use the super awesome code he wrote, go ahead, but leave me out of it unless you can get him to complete what he started and not leave us with a half-finished demo which can't be made real, we're not supporting the shit he leaves behind him".
Everyone else considered him a nightmare, because while he was prolific and wrote cool looking things ... it was all smoke and mirrors, which covered about 2% of the functionality and 75% of the sales demo.
I was glad when I transferred to another group and he was someone else's problem. Though, I did notice the carnage of sales people and clients who bought into the bullshit and thought they were getting something real only to find it had an awful lot of "pay no attention to the man behind the curtain".
I hope he was a very unusual case, but he was an absolutely prolific coder, who wrote absolutely terrible code.