Learning from mistakes

Sometimes I will try out a new technique to see if I can improve on the art in Overgrowth, or make something just as good but in less time. One new technique I was thinking about was applying a very common technique used for making 3D buildings in games to models of natural terrain. Here are some textures for it:

What is interesting about the textures is that the main texture on the left is made up of smaller tiling sub-textures (the letters designate the different sub-texture areas). I have used these textures to cover a larger area on the right, which has seamless transitions:

I thought this was looking okay until I tried to actually make something. The results were underwhelming:

I felt it looked too cartoony, and I didn't like how the stone texture came out, so I tried the same thing over again with different textures. It came out even worse:

Even though I was unhappy with these, and we won't be using them in Overgrowth, it was nice to experiment. This technique would work great for a game that was a bit more cartoony. The basic idea for the technique is still pretty useful so now I am working on applying it to trees. So far I am pretty happy with the progress even though it still has some obvious stretching. With some tweaks I think it will look decent, and it is easier than what I was doing before:

I feel that failure is a natural part of learning and exploring new ideas. Even when things look bad it is kind of fun and interesting. How do you deal with failure?