This article may make you wonder why to use Python at all -- as it says in the conclusion, Python may never rise to the level of popularity of Perl. I would like to point out that Python particularly shines in the area of numerical analysis. Python is heavily used for numerical work in physics and engineering, and since recently also in bioinformatics, particularly for gene expression data analysis. The Numerical Python package (numpy) has become the de facto standard for numerical analysis with Python, and provides many mathematical functions that would have to be written from scratch otherwise. In addition, as mentioned in the article, the tight integration between Python and C makes it possible to implement numerically intensive routines in C, and call them from Python. This then gives a very flexible environment for data analysis. As far as I know, Perl is not very suitable for numerics.
I would further like to point out that using PyCrust might make things more difficult for beginning users, especially because of the installation process. Instead, you can use IDLE, the standard Python shell which is part of the generic Python distribution. It doesn't require any additional installation steps, and should run on any platform. PyCrust may provide some features that IDLE doesn't provide. I have used IDLE for several years though, and have been quite happy with it.
Michiel de Hoon
University of Tokyo, Human Genome Center