I am about finishing its nose part already. Say, I am at 20%.
As for the method - it is believed (by myself) that this way I could achieve extra coherence between the model and the texture.
The way we all do UV mapping here is tile-based, whose one of strength is unpredictability of the final result (which makes things a little more interesting).
Such an iterative process is perfect for creators like me, who do not see the image completely at the beginning, but have to mess around to start seeing details and peeking the design out of them.
Modeling > texturing > modeling (and maybe texturing again?) is a nice way to make a good use of predefined textures and basic shape... Though I am still grasping around the whole concept of this technique (for future publication and tutorial-recording).