In Fig.09 you can see just the dirt map on the left and how the Offset and Tiling coordinates have been modified to position the map along the right hand edge. It is tiled only along the U axis, which corresponds to the length of the corridor.
The Tile checkbox is not ticked in the V axis and therefore only partially covers the width. The dirt texture (Brown08) is tiled by 6.0 in both axes and so covers the entire floor area, but the mask reduces its visibility to just the right edge within the white area.
Layer 4 in the composite map represents the line of dirt along the opposite wall, the mask of which is visible in Fig.10. As in the previous layer, the mask has been offset and tiled to only run along the wall edge whereas the dirt texture covers the entire floor. The difference here is that the dirt texture is set to Multiply as opposed to Normal.
When activated, the result can be seen in Fig.11.
When this layer is now made visible, along with the previous three, we can clearly see how the composite map works compared to Fig.05 where only the base layer is evident (Fig.12).
When all eight layers are active the result is a muti-textured surface that has avoided any unwrapping or custom texture creation in Photoshop (Fig.13).
The advantage of using this technique is that it only requires a simple planar map and dispenses with the need for unwrapping as the coordinates for each of the textures can be manipulated individually in order to localise detail, using a mask to control what is visible.
The trade off is that the scene can be heavily laden with textures as a single surface can use several as part of a Composite map, so depending on the project limitations this should be a consideration.