Not sure how to work around file dependencies in generative design (Revit2023)

We tested the movement of a generative design (Revit 2023) that included file dependencies.

The following four items were tested for comparison

➀File dependencies: one Excel file (two sheets within)
Two csv files
(2) Generative design “Include dependencies”: Unchecked
Checked
(3) Menu to generate studies: Dynamo menu
Revit>Manage menu
➃Action when file dependencies are replaced

The results can be summarized as follows. (This is only the result of this experiment)

nly the following combinations returned complete results.
Referenced files : 2csv files
“Include dependencies”: checked
Generate a study from “Dynamo”

When the file dependencies are two sheets of one Excel file, the connected components are seen as one, and the design context parsing does not work well(?). The results are not stable.
Regardless of whether dependencies are included or not, a files folder is created in the Dependencies folder at an unknown time, and file dependencies are copied into it, but whether they are generated using this folder or not is not clear.
It is not clear if the dyn file f is saved with the dependencies included, but it may be more stable to generate studies from Dynamo’s generative design than from Revit’s generative design. (I’m not sure, memory pressure may be slightly different.)
When the file dependencies were suddenly replaced separately and made to work,
Only the following combinations returned complete results.
File dependencies: 2 csv files
Include dependencies": checked
Study generation is started from Dynamo

What does “include dependent files” do? I had no information, so I asked GPT and received the following plausible answer.
>Including file dependencies when generating a design allows the software to analyze and optimize the entire design context, taking into account all interconnected components and their relationships.

This comprehensive analysis ensures that the generated design is coherent and compatible with all linked files, resulting in a more accurate representation of the final product.

I would like to address the following questions

  1. Is the “Include Dependency Files” checked status by itself able to change the results of the analysis?
    Or does it have meaning when saved as dyn for GnerativeDesign? (If the latter, I think the menu position is different.)

  2. When the “Include file dependencies” is checked, when are the referenced files included in the Dependancies folder?
    (When the file is included, is the entire design context analyzed and optimized?)

  3. With the “Include Dependencies” checked, are the saved dyn files for Generative Design saved with the “Include Dependencies” check (when they are imported, have the entire design context been analyzed and optimized)?
    In other words, does the same analysis occur when generating a study from Dynamo and when generating a study from Revit, or is it different?