Best way to optimize problem with additively separable fitness function?

I am using a genetic algorithm to maximize a few hundred thousand real-valued variables. Each of the variables, $x_i$, has its own independent boundary condition.

The fitness function uses each of these variables to compute another value and returns the sum of everything:

$$fitness = g(x_1) + g(x_2) + g(x_3) \ + \ ...$$

This is taking incredibly long in python. In this situation, what do I gain by maximizing all values at the same time, i.e. using the genetic algorithm to find the global optimum in this huge search space? Does it even make sense or should I just maximize each variable independently, since the fitness function is additively separable?

2nd question: If I include some constraints which depend on the value of other additively separable functions, like:

$$\text{if } h(x_1) + h(x_2) + h(x_3) \ + \ ... 0:$$ $$fitness = fitness + penalty,$$

can I still maximize each variable independently in any way? What is the best approach in this case?

Edit: Changed optimize to maximize to make things clearer.