Sky-sphere lighting is NOT global illumination. Global illumination (or GI for short) reffers to the globabl interaction of light among all objects. Radiosity is an example of a limited kind of globabl illumination. More complete models of global illumination are present in renderers such as Lightflow (with caustics, soft-shadows, reflections and refractions, diffuse light interaction, etc.).
Now, granted, this is simply an issue of termanology. However, termanology can be very important, since it is what allows you to communicate with other people effectively. Thus, I figured I might as well point out that you are using termonology–which is already in place–incorrectly.
Another example of mis-used termonology is the term “radiosity”. Radiosity, it turns out, is NOT the “color bleeding” effect. Rather, it is a very specific way of calculating that color bleeding, by dividing the scene into many polygons. The color bleeding effect is actually reffered to as “diffuse light interaction”.
As many of you may have noticed, I have used the term “radiosity” incorrectly myself. However, once I learned that I was using it incorrectly, I stopped doing so. Please do the same with “GI”.
As a side note, another much easier way to achieve sky-sphere, or “hemisphere”, lighting is to use the “hemi” light source type. Unluckily, this will not calculate shadows or specularity correctly. But it does calculate the diffuse lighting correctly. The specular lighting really isn’t all that important (it still looks fine), and the shadows can be faked by using a very blurry spot-light.