Good article but I'm not sure the thinking is complete (I'd have to think about it!). There's a big hole, MLB-quality pitchers are mentioned therefore the first baseline graph isn't appropriate.
Regarding your musing about whether it applies to programmers, don't forget the curve depends alot on whether they're being "pitched" by MLB-quality problems!
(There are plenty of good coders whose .401 average is because they're fiddling with less than intricate business processes).