It's neither. The most important thing is to produce results in as little time as possible. You don't even know if you're solving the right problem most of the

It's neither. The most important thing is to produce results in as little time as possible. You don't even know if you're solving the right problem most of the time. So you want to better understand what's the problem you should be solving before you spend more time trying to solve it. The second most important thing might be to become able to recognize when it's the right time to change/rewrite code to be of better quality. Do it too soon and you're wasting time. Do it too late and it takes more time.