Neat.
Something of this size, so small (well big in this case, but small pixel dimensions) needs to use each pixel carefully. Consider what each pixel is doing to carve the shape of his face. The relationship between neighboring pixels is very important. Some areas, like the nose for example, aren't rendered out properly. You have to really put the right pixel in the right spot to define the shape correctly.
When you've got groups of pixels in different colors like this | | | | you are missing an opportunity to make things a higher resolution. Does that make sense?