Thanks for this update !
An other question about .rgirlsx files : If I completely delete some .rgirlsx files, will the girls generated with those files still "work" ?
I'm asking this because I might try this in order to get some girls more frequently than others, so deleting the files for a while (after saving them somewhere else) could do that, but I'd still like the girls previously generated with those files to "work".
An other way to cheat the frequency of seeing some girls might be to have duplicate files for the same girls, just changing the name of the file (adding 2, 3, 4, 5, etc... at the end of the file). Do you think it would work ?
So I did the following experiments:
1) Created a copy of "Aaliyah Love.girlsx" and started a new game. Saved it immediately without doing anything. In the 'list of girl files' at the beginning of the save file, there are two entries for Aaliyah Love, and the stats for Aaliyah Love are saved twice.
2) Deleted the copy of "Aaliyah Love - Copy.girlsx". Restarted the game, saved immediately. There was no apparent change to the save file. Both the deleted filename and the two copies of Aaliyah Love remained.
3) Without changing any files, restarted the game, ran a turn, saved the game again. There was no apparent change to the save file. Both the deleted filename and the two copies of Aaliyah Love remained.
4) Created a copy of "Aaliyah Love - Copy.girlsx". (Same as the original name). Ran a turn. No apparent change to save file.
5) Created a copy "Aaliyah Love - Copy (2).girlsx". Saved the game without doing anything. There were three filenames in the file list, and three copies of Aaliyah Love in the girls section.
So it looks like your last option will work, at least for .girlsx files.
I've done no experimentation with .rgirlsx files. They are not kept in a file list that I can find in the save file. Random girls that get turned into an in-game girl are entered in the girls section. My suspicion is that the .rgirlsx files are dynamic for every run, and that the same trick will work - multiple copies of the same random girl file should improve her odds. But I don't feel like running the trials and the statistical analysis that would be required to verify this hypothesis. Ask a coder, they can tell without experimentation.