I think it would be really easy to do. You just need to edit a couple of text files, and then move files into the appropriate directory.
You can already just overwrite files with different sounds of the same name to modify kits on your computer. That’s about as easy an interface as I can imagine.
In this thread, Brad Holland indicated that they are willing / able to share the format documentation, but that hasn’t happened just yet. I think that would only take an hour or two.
I think experimentation will yield the answers. It’s not cryptographically sealed or anything.
The .KIT files are dense gibberish to me, but the .S2400 files show some obvious structure (it is sample/bank based, and “0” is the null value).
Anyway, you could call some sort of script to modify these files or just do it by hand. Batch processing of audio files is pretty straightforward as well… use macros in reaper or audacity.