All of them seem to cover the basic functionalities needed to construct phylogenetic trees. However, does anyone have some recommendations or can name some distinct features that should point me to using one of the above mentioned over another?
In phylogenetics, you can not rely on a single methods because each method has its strength and weakness. For example, Maximum parsimony tend to be more sensitive to long branch attraction. The choice of method depends a lot on the kind of research question you want to address and which kind of data are accessible. In short, it is important to know all the different methods including the limitations and chose the analyses working best for your dataset. It exist a lot of different software and most of them perform basic analyses rather well but it gets a bit more important to chose the best available analyticla method and the best software if your research problem is a bit more complex.
PAUP hasn't been updated in many years and is excruciatingly slow for any large dataset. PhyML will do most thing PAUP does much faster in a Likelihood framework. MrBayes is usually quite fast and operates on a Bayesian framework which is useful to enter any additional information you might have on your taxonomic set.
In phylogenetics, you can not rely on a single methods because each method has its strength and weakness. For example, Maximum parsimony tend to be more sensitive to long branch attraction. The choice of method depends a lot on the kind of research question you want to address and which kind of data are accessible. In short, it is important to know all the different methods including the limitations and chose the analyses working best for your dataset. It exist a lot of different software and most of them perform basic analyses rather well but it gets a bit more important to chose the best available analyticla method and the best software if your research problem is a bit more complex.