Importing non-duplicate records only

I have tried to follow the example Fenton Jones gave (http://www.fentonjones.com/Articles/10_Import_Dupes.html) for importing new records and avoiding duplicates, but the intermediate script does not make sense nor does it act as desired.

I am trying to filter out duplicates in one set of temporary records in a .fp7 database for import into another .fp7 database.

Fenton's instructions were as follows:

Find All

i.e. Show All Records in FMP 9

Go To Layout ["Concat Only"]

layout with single concatenated field of First and Last Name only

Copy All Records

Go To Layout ["Concat g"]

layout with global concatenated field

Paste [select, No style, "Concat g"]

Exit Record/Request

i.e. Commit Records in FMP 9

Go To Related Record [show, "Concat g to Concat"

GTRR in temporary database

Perform Script [External: "Transit"]

a two-step script:

Delete All Records

Find All

"The first part of the script copies the Concat field from all records, then pastes them all into one global field, Concat g, separated by returns. The last 3 steps will go to only the records in Transit that are duplicates of ones in Import, using the many-to-many relationship, then delete them all. The last step will find the remaining records, which could be imported into Import."

My trouble is at the GTRR step. Working in FMP 9, the script steps became:

Go To Related Record [show only related records; From table: "Transit"; External; Using layout: "Transit" (Transit); New window]

My script/relationship ends up finding all records in the external file instead of just the ones already in the working (Import) table (i.e. those matching any of the global field Concat g entries). As a result, even the new unique records are deleted in the last step, leaving nothing to import.

I appreciate any and all assistance. Thank you!