Here is a sample Scrivener project that demonstrates the workflow:
IndexV1.11.scriv.zip (574.4 KB)
I’ve added a variable doTitleCase in the script to enable/disable the title case changes, redownload the script V1.11, set it to false to stop case changes.
For Word, the way it shows index terms is expected, the {XE:Term} is how it shows the index fields, if you don’t want to see it just turn off whitespace visualisation:
For ancestors, you use colons, so this in Scrivener:
…becomes this in Word:
