One approach is to write to entertain the game master so reading through the material is enjoyable on its own. There is a certain truth to that. The designer needs to create an emotional response from the game master reading the work. Without the game master getting excited to use the material, it will remain in the book and never see play. The game master needs to see in their mind the possibility of a great game session in their mind before that session actually happens. If not, they will not want to use the designer’s work. Also, many people just like collecting and reading game books. I have a shelf full of stuff I’ve never run or have only run once or twice. Reading the rules and setting material can be entertaining and enlightening on its own. I’ve learned a lot reading the Amber Diceless Roleplaying book but I haven’t played it since the mid 90’s.
On the other hand, there is the pure utility approach. Why is the dungeon here? Where did the monsters come from? Why are the aliens abducting the humans and probing them? Who cares, here is the stat block for the aliens, a map and a key for their ship. Go! The utility first approach makes using a tool at the table far easier and helps the pacing of the game. I hate having to take 30 seconds to read through a room description to figure out what the players see and what is hidden. Even worse, if the prose is too descriptive, you have to work through it to figure out what is important and what isn’t. It slows everything down and breaks the immersive quality of the game I’m trying to accomplish. Taken to the extreme, this method can feel emotionally flat when you’re reading it. The adventure or location becomes a map and a set of plain stat blocks. This was common in some of the early original D&D adventures. The room description would be something like, “1. 3 goblins” and that’s it. Functional, easy to understand but not very exciting..
I think the way to go is a blend. Parts of an RPG text can and should be entertaining for the game master. Catalyzing their excitement is important. If the game master isn’t excited to run the material then it won’t be run at all or it won’t be presented in a way that the players will enjoy either. The parts of the text that the game master will need to access during play should be as simple as possible. It should not require extensive note taking or look ups and cross references. Perhaps a quick reference sheet for the game master to keep the session moving without the need to make their own. The OSR has been working on this particular problem and we are starting to see some workable formats come out of the scene. I hope that some of the solutions get adopted by the wider RPG industry. It will make running games a lot easier and prep more enjoyable for game masters.