I think there are some details missing here in your description. Normally, use values from field value lists draw their values from indexed fields. It is indeed possible to draw values from an unindexed field if at least one of the two fields in a "use values from field" value list IS indexed and is the "sorted" field of the two. But in such cases, the data from each unindexed field becomes a single value in the list even when you have return separated values in the field. For a global field, this produces a value list of a single value.
Are you sure that your global field is not serving as a match field in a relationship to filter the values of your value list?
Thank you for the response, I apologize if my explanation is unclear, it's been a while since I set it up.
The value list is based off of the relationship in the picture. The actual fields don't matter, and the relationship doesn't matter, but somehow the relationship enables filemaker to use an unindexed field for a value list. I'm not sure if I am allowed to post external links, but a more in-depth implementation of the technique and a demo are located here: http://filemakerhacks.com/2012/07/25/magic-value-lists/
The ExecuteSQL function to populate the list is also quite simple, after determining the context for the valuelist I use Set Field ( ExecuteSQL ( "select Datapoint FROM DataTable WHERE FieldID = ? and ForeignKey = ?" ; "" ; "" ; $$CurrentFieldID ; $$CurrentForeignKey )) to set the global field, commit records, then re-enter the field and viola, the valuelist populates correctly.
So at the end of the day, as odd as it sounds, it works. My only issue is in the indexing of this Valuelist_Source field, and getting Auto-complete to work with the resulting value list.