Optimisation strategies from Darwinian competition?

Background

Simple Selection from Martin Nowaks book Evolutionary Dynamics, or ‘Survival of the fittest’ when total population is held constant

We define x_i to be the size of the population of the i-th type(phenotype).

We define f_i to be the fertility of the i-th type. A very simplistic model where the population size is not held constant, would then be:

\dot{x}_i = f_i x_i

We now define the population size across all types to be constant; \sum\dot{x}_i=0. To account for this extra definition, we extend the simplistic model to:

\dot{x}_i=x_i  ( f_i - \phi )

Then \phi\sum x_i=\sum f_i x_i. If x_i is the respective type frequency in the population, we can make a further simplification that \sum x_i=1. Which gives the standard formula for selection:

\phi=\sum f_i x_i

The variable \phi can be thought of as an average fitness. Then it is easily seen that the type with the largest f_i always wins; or in a phrase ‘survival of the fittest’. Criticism of phrase here please.