Why I Write Concept Software

I'm somewhat puzzled how a designer could, when confronted with the task of creating a real-life product with real-life constraints, do so with compelling results if they had not already spent time thinking about assorted what-ifs involving various self-imposed constraints.

Are there good designers (and for that matter, good developers) who do not spend time sketching out assorted concept whatevers to play with ideas and stretch their imagination?

Often, while working on a less than thrilling project, I'll come across some chunk of code that works and solves an immediate problem, but gets me wondering if there's some other way to accomplish the same thing. Often, this sort of thing gets tagged as "too clever" coding, and often that's exactly what it is. But these exercises lead to a better understanding of how to use this or that language, or how to think about this or that problem, so that later on, when I'm solving a real problem, I have more ideas to draw from.