In the 19th century there was a debate on orchestration versus construction. When you compose a symphony orchestration is very important. If you put the pair the wrong instruments your symphony can sound horrible. Some composers decided to make the symphony mostly about orchestration and others were all about the structure of the composition. During the process of both you have to weigh the positives and negatives of sacrificing one or the other.
If you wanted an orchestra to sound really good you would have to sacrifice some of the structure. Likewise if you wanted great structure you would have to sacrifice orchestration. The big debate was between the composers Tchaikovsky and Brahms. Tchaikovsky' symphonies sounded great but lacked the structural components that Brahms had. Brahms was the reverse. Naturally you want your orchestra to both have amazing structure and be well orchestrated. However, that is not always possible.
So if we flash forward to today. I find the exact same argument among game developers. Does the game play well or look awesome? Do you sacrifice definition for mechanics? Naturally you want a game's design to flow and in a perfect world you would have great mechanics and great assets. But a long the way you have to sacrifice something. The question is what will you sacrifice.
Personally, construction will always trump polish. To me all forms of media should be engaging and I consume media for the ideas and the exploration. On top of this, well constructed ideas can be adapted and improved upon. The oposite to this is fluff. Games that look and feel good but have no substance. It's the classic style versus substance. I like my games to have substance.