Many application languages have "reflection" API's that can examine internal or low-level code structure. I used database languages such as dBASE (and clones) that are based on more or less on sequential query languages. While I did like more control over the intermediate steps, including the ability to analyze them; the problem is that different people do things too differently. SQL reigns in the "creativity" to a large extent. What works well for an individual may not scale to "team code". Working with a Picasso-coder's code can be a bear.

This is a valid argument. A high level abstraction definitely serves as a restriction over code that gets to "creative" at the expense of obfuscation. Following this line of logic, the argument then truly does become apples to apples.