Resolve conflict with weight windows and global russian roulette

[GuySten]

Description

Currently the global neutron russian roulette (which is used when survival biasing is turned on) conflicts with the application of weight windows.
As suggested in #2773 (comment), this PR makes global neutron russian roulette apply only when neutrons are outside of the weight windows mesh boundaries.

Fixes #2773

Checklist

• I have performed a self-review of my own code
• I have run clang-format (version 15) on any C++ source files (if applicable)
• I have followed the style guidelines for Python source files (if applicable)
• I have made corresponding changes to the documentation (if applicable)
• I have added tests that prove my fix is effective or that my feature works (if applicable)

[pshriwise]

Thanks @GuySten! You are a machine. Some thoughts to consider here in the comments.

[GuySten]

I've restructured the code to be more backwards compatible.

[GuySten]

I've further improved the code structure according to your suggestion.
I will have to dig into why the tests fail.

[GuySten]

This PR is now ready for another review iteration.