The guidelines seem to treat styles differently. Why is that?

Some styles are quite well-known, others are historical notions, while still others are artificial creations for the purpose of categorizing relatively unique beers or for grouping similar beers for judging purposes. That said, there is a notion of “narrowness” of style that applies to the variation between commercial examples within a style. Some styles are based on a small number of examples (e.g., California Common), while others may have explicit requirements (e.g., Kölsch). These are narrow styles. Some styles embrace multiple stylistic variations (e.g., Foreign Style Stout, Old Ale), so are broader. Some styles allow a great degree of creativity on the part of the brewer, and therefore are wide open (e.g., Mild, Belgian Dark Strong Ale). All of these factors contribute to styles being handled differently.

The nature of the research into the styles is another factor. Some styles are quite well-known and have many commercial examples. These styles are relatively easy to describe. Some styles are historical, have few sources, or are not widely available. These styles may be less fully described. Styles also tend to evolve, and descriptions may describe variations over time. In some cases (e.g., English IPA), the styles describe beers the way they used to be made more so than the way they are currently made. This allows the historic heritage of a style to be preserved and brewed by homebrewers, even if commercial brewers no longer make them that way. Styles may be rediscovered (e.g., Porter, Witbier) and be revived in their historical context. It is a judgment call on the part of the BJCP to decide how best to handle a style. We tend to describe beers in the way that they were when they were the most authentic and popular, although there are exceptions.