One feature of the paper that will be useful is the structure. First a overview of the string method and past implementations and advances is provided. Next a theoretical framework is developed; from the physics of diffusion to the definition of the pathway. Once the framework is established the authors move onto the dynamics of the reaction pathway. A few paragraphs are spent discussing the committor function (the probability of reaching state B before state A). Lastly, some issues with implementation are discussed.
Another presentation method I will emulate is the authors' presentation of figures. All of the explaination necessary to understand the figure is located directly below each plot. This removes the need to hunt through the paper in search of pointers to the figure.
One improvement I would make to the paper is to include empirical evidence as a baseline for comparison of the various methods.