Do Tasks Quickly and Imperfectly

Why don't you get stuff out, determine what needs to be improved (if any, as opposed to starting a new task) and improve that?

If you put out a marketing video for your app, and it drives 1000 purchases of your app, you should be proud of that. The video itself is irrelevant, so don't judge it, or yourself by it.

I think the craftsman analogy is misleading because physical objects can't be updated after selling. Maybe a better analogy is that of a chef, who doesn't measure himself by how it turned out when he tried a new dish for the first time. He measures himself by how it is after repeated iterations.

As Basecamp says, work should always be compared with what it was before, not a theoretical ideal.