Not necessarily right. you can try to use 2*2 mesh and check the result. FDE solves eighgen values so it needs some details of the structure.
The most important thing is: when Metal BC (the default BC for mode expansion monitor) is used it has box modes, which are not necessarily the propagation modes, whereas using PML BCs can have a lot of PML modes that need to e removed. Since you set a certain number of tiral modes, when the solver reaches this number it will stop, so some higher order modes may not be found.
You are navigating away from the AIS Discovery experience