I did it again, and it looks better. So is it true that the more polygons an object has the better the 'folding' and 'bending' of the mesh will look?
Ok, Zoo was faster and already shows the subdiv modifier which by the way is nothing else than the computer creating even more geo, but smoothed. Subdiv modeling needs a bit learning and exercise, though, and understanding of topology.
On the other hand you already have more geo than you really need imho (on the other hand there is a discussion about it if you should use holding edges or not which are necessesary when you use less geo). You need the geo in the right places. I strongly recommend to watch some tutorials. Some are here, some others you'll find on youtube, vimeo etc. They don't have to use Cheetah if it's just about understanding what to do (should be a subdiv modeler. If you know your way in Cheetah you can watch tuts with maya, 3ds max, silo, blender, modo, whatever: It's in this case not about redo what others did but what to do (the problem is, all those apps for sure use some tools that don't exist in Cheetah. One has to translate). After a while you'll be able to discern which are good ones and which are not so helpful.
And it doesn't matter how much I'll learn, I always watch some tutorials in my spare time to see how others work, how others tackle topology problems etc. With 3d, at least for most of us, the learning process never stops.