Does studying computational electromagnetics have any application beyond writing CEM software like HFSS?

[shredder929]

Computational Electromagnetics seems like a really interesting field, but the posts I've seen say it's not worth studying except to understand what's going on under the hood of your simulation software.

Is this true, or do EM engineers frequently make use of knowing FDTD methods to tune it for their purposes? Is it a viable specialization in its own right?

 

[BradtheRad]

Similar question, how much can we learn from studying program algorithms inside plain electronic circuit simulators?

Our best use of the tool is for us to learn more advanced things by it, and achieve bigger projects by it.

For a minute I thought this question was about optical computing. It promises huge advances in speed and miniaturization.
No current return paths. Reduced power needs.
No magnetic fields emanating unexpectedly and causing interference.

Now, suppose someone out there is knowledgeable about both electromagnetic propagation and how HFSS calculates with radio waves? Regardless that light acts differently from radio, it could lead to a breakthrough in optical computing...