It works that way because we have encoded agreed upon negatives into law and/or regulations. By contrast, the positives are still highly subjective and not broadly agreed upon.

So the actual process is dominated by "Does the project avoid X?", "Does the project ensure that Y cannot happen?", "Will the project ever exceed limit Z?" rather than "Does the project bring <positive-good> to its location?", because people have not yet agreed upon a clear definition of <positive-good> things. In contrast, the stuff to avoid makes itself known, and the stuff we can mostly agree on becomes part of our civic codes.