Think of the way normal images are handled.

You have 3 (or 4) seperate channels in your image. Red, Green, Blue and sometimes Alpha. In order to create a specific image each channel is filled in Grey scale with black being the completely without any of that color and white being completely whatever color the channel represents. By blending these various channels you create the actual image that you see on your computer.

When you apply this to shaders - each channel represents not a color...but a seperate picture. By using a base image (your large terrain skin) you are able to make much more complicated skins. You can specify the Red (R channel) to be a stone texture, the Green (G channel) to be a grass texture and the Blue (B channel) to be a water texture. By doing as much and mixing the channels using the various shades of grey in each channel through a shader you are able to display grass or moss covered stone or swampy grass. You can also have areas that are completely grass, water or stone.

It really all depends on the shader used (you can right a shader to have different results when overlaying textures in this manner similiar to the various options in Photoshop like color dodge overlay, hard light overlay or other effects). In this case, the shader acts more like adjusting the transparency of the layers and simply adds the pixels together based on the value of their assigned channel.