The release notes just mention "Improved Export to 24p".

Could someone elaborate on what exactly has been improved and how?

24p export

Previous support for 24p did not work on all camera types, or required the user to manually input a setting to match their particular camera type.

The new functionality should work for all camera types without any user input, other than turning 24p processing on.

The old system was detecting which camera was used to take the footage, the new algorithm compares video frames and looks for duplicates or near duplicates to detect the pull down pattern, then processes this accordingly.