Apologies if this question it too open-ended; evolutionary biology is not my primary field.

I have been reading a lot about the use of statistical mechanics in analyzing evolutionary dynamics. As an example, consider this paper: http://www.pnas.org/content/102/27/9541.short.

The ideas are pretty straightforward. There is some genotype-fitness landscape and selection drives populations towards maximal fitness. However, genetic drift can pull these populations away from maximal fitness. These two concepts are analogous to minimization of internal energy and maximization of entropy in thermodynamics.

However, most of the papers that develop these models never properly explain how to apply them to observable phenomena. Is it because there is some disconnect between the parameters in the models and nature? For example, many of these models use the effective population size, which is unknown for most species.

I guess my main question is this: are there any demonstrations on how to use these toy models to predict something in nature, or are they simply for theoretical purposes.