The Radar Cross Section (RCS) of electrically large targets including cavities (such as jet engine inlets) remains a challenging problem in electromagnetic modeling. Rigorous methods are known to provide a good accuracy, since only the approximation due to meshing is considered. Nevertheless, these methods are difficult to use for targets much greater than the wavelength due to the computation time and memory requirements. For electrically large targets, High Frequency asymptotic methods can be used, among which Iterative Physical Optics (IPO) is very efficient. IPO is an asymptotic numerical method based on Physical Optics (PO). This method, first introduced in 1995 by Obelleiro et al, is known to provide a good trade-off between complexity and accuracy. Since 1995, several improvements were published. More recently, some improvements were made to the IPO method by the help of fast numerical techniques, generally inherited from advances made for rigorous methods. Given all these previous studies, the IPO method can be implemented into very various forms depending on the desired application. Thus, the goal of this paper is to provide a complete overview of the recent improvements on Iterative Physical Optics. The advantages and drawbacks of each specific form and acceleration of IPO will be discussed in order to give some advices on the choice of the best IPO form for a given problem.