These weren't un-executed ideas, though. I think it lines up very well with my general sense that the best way to get to a really good implementation of something is to publish one that's lacking in some way that annoys people. Then you have an idea plus something people can contribute work to improve.