It's very easy to mix correlation with causality in such studies. Being a sailor or pilot is not hereditary. A lot of abilities that make one excel at sea or in the air, however, are.

A (professionally skewed and quite materialistic) interpretation of the DNA story - our (and every living being's) DNA acts as a collection of algorithms. The input against which our 'algorithms' are tested directly or indirectly is our very own environment. Some of the implementing code is new, some ancient and patchy, maybe redundant, some more robust, some is buggy and error prone. Just as we know there is no 'perfect' algorithm in computer sciences, there is no perfect DNA in nature. Some algorithms work only with certain premises, some are restrictive in that regard, some are not, some are quicker, some slower, more or less complex, etc, etc. However, you can only compare each only to ANOTHER one (under varying datasets -> environmental conditions) but there is no 'absolute', maybe another, vastly superior one is just around the corner.