The one thing that took me a long time to realize is that even when it comes to something like software if you throw enough half-assed labor at a problem there is a reasonable chance that eventually something of significant legitimate value can be created. No, it probably won't be as good as something created by a bunch of geniuses, but it can still end up being "good". When you play the long game or the big game you can afford to rely on half-assedness to work out in the end.