The area is split into four quadrants. On the X axis we have the abstract vs the concrete. Moving along that line to the left we get more vague and less detailed concepts being communicated on the tabletop and more detailed (less abstract) ones being communicated as we move to the right. On the Y axis we have realism, more realism up, less realism down. By realism I mean plausibility, not if the event can occur in 21st century Earth, but rather if within the context of the game the event seems "right", if the outcome of an action seems coherent with player expectations and is "intuitive". It also has to do with how fluid the game feels, we don't live life in "stop motion", to us life is fluid and continuous. A game that is more fluid and stops less to check things imply feels more realistic than a choppy one in which game mechanics resolution consumes a great deal of the time.
"Simulationist" on this graph is not a position, but a means to go from the abstract to the concrete. It is a process that fills in detail lost by the abstraction process. This can also be achieved by another process called storytelling. Simulationist is a process to fill in detail by which rules and dice mechanics come into play to "unbias" the GM or player by injecting randomness and binding things to certain constraints. How well crafted these dice mechanics and rules are made influence in the player experience. The storytelling process on the other hand is less bound to randomness and so players and GMs may fall into certain habits and be biased, the are also less bound to certain constraints and free to improvise on the detail. This in turn can create uncertainty with the players and be counter intuitive. The GM may pull out some unexpected outcome and break "realism".
It is important to note that neither "simulationist" nor "storytelling" add realism to a game, they add detail. If the detail added seems unreal, out of place or causes a choppy flow of the game it is not making the game more real, it is simply making the player experiences more detailed.
There are two types of concrete to consider. The one graphed on the diagram is concrete is the communication between the players. Inside the player's mind the process of going from the abstract to the concrete continues, as a player you begin to fill in more and more detail to recreate the event in your mind. For example an abstract statement from the GM like "The goblin swings at you and does 5 points of damage" will be reconstructed in your mind to a vision of the goblin sword your character's shield and armor, the place, the movements, swings, etc. A more concrete statement from the GM will give you less leeway as to what your mind can do, it narrows down your options as you reconstruct the scene in your mind. Less detail on the graph (games on the left) means more imagination leeway for you, more detail on the graph (games on the right) means less imagination leeway.
Up and down on the graph has to do with how well the communication process facilitates this reconstruction in the players mind. If the communication process begins to fall into the unreal area it will begin to cause issues with the player's reconstruction process. "Glitches" in this communication may make it less intuitive for the player to create a convincing reconstruction based on what's been communicated. For example a character attacks and hits the goblin, but then suddenly falls victim to an unexpected attack of opportunity. Although it all falls well within the rules the game, the player will be building this image of the attack in his o her mind and suddenly the attack of opportunity comes along and the player goes "hey what was that?". The attack is an unexpected and counter intuitive event that sets the player's mind off balance. It hinders the abstract con concrete process that goes in our minds when we are communicating.
This takes me to ask what is a good game and what is a bad game? There's a lot of opinions on this and some will comment that there is no such thing as an objectively better or good game. I do disagree with that. Independent of genre, simulationist or not, there are game properties held by some which make them better than others, that is objectively better. I believe a good measure of a game is how well the rules create a communication process between players which in turn facilitate the recreation process in the recipient's mind. Games on the bottom of the graph, quadrants 3 & 4 are not as good as games on quadrants 1 & 2. Those on quadrants 1 & 2 offer the same level of detail than their lower quadrant counterparts, but facilitate the reconstruction process by lacking inconsistencies that may hinder the player's mind from working up more detail. It doesn't matter if the rules offer more or less detail, in the minds of the players the abstraction of the game is reconstituted into a concrete experience much faster and easier with those above the X axis than those below the X axis.