I'm unfamiliar with the buck converter model. But I can show you a simple example below to demonstrate the basic model-based design concept. That's why having the model is generally very useful in design. So, if you have a Simulink model (deterministic system), the math model 100% known, or can be obtained from analyzing the mathematics of every single block, in theory.
Example:
Say the transfer function of the linear system (plant) is known. Since it is a first-order system, you can try a PI controller 
. Now, the closed-loop transfer function with a unity feedback is given by
.By equating to zero the denominator of a transfer function, you get the Characteristic Equation:
,where the roots (aka poles) of the characteristic equation must have negative real parts for the closed-loop system 
to be stable. If you already have a desired Characteristic Equation:
.Then you can compare the terms and solve both equations directly to obtain proportional and integral gains at:
, and 
.This design method is called Pole Placement, because you select the values for 
and 
such that the closed-loop poles are placed to specified locations, in this case at, 
, because we know that poles with negative real parts ensure system stability. If you have the Control System Toolbox installed, then you can also use the place() command. 
using high school algebra (i.e., solving simultaneous equations) and then, to simulate whether it can satisfy the performance requirements or not.
