Hi Aries S. , not sure if this is posted in the right forum (Developers), but I'll try answer here as best I can.  Since you have the partially pre-translated XLIFF files, there are ways to salvage your work and minimize the need to reprocess everything from scratch.

Your Initial Approach
Recreating the project and using Perfect Match with the XLIFF files from the initial run is a valid method. Here's how you can proceed:

1. Recreate the Project:
   - Start by creating a new project in SDL Trados Studio with the original source files.

2. Use Perfect Match during Project Preparation:
   - During the preparation phase, use the "Perfect Match" feature to leverage the translations from your partially translated XLIFF files.
   - To do this, go to Batch Tasks > Perfect Match, and select your XLIFF files as the reference files.
   - This will apply the existing translations to the corresponding segments in your new project.

3. Run DeepL Plugin on Untranslated Segments:
   - After applying Perfect Match, use the DeepL plugin to translate any remaining untranslated segments.
   - This ensures that only segments without existing translations are sent to DeepL, saving time and potential costs.

Considerations with Perfect Match:

- Segment Status and Attributes:
  - Perfect Match preserves segment statuses and attributes from the reference files.
  - If the segments from the initial run were marked as "Translated" or had specific statuses, these will carry over.
  - If you prefer not to mark these segments as Perfect Matches (e.g., to avoid confusion in the review stage), you might need to adjust the segment statuses after the Perfect Match process.

- Workflow Impact:
  - Be aware of how Perfect Match segments are treated in your workflow. Some QA checks or automated processes might handle Perfect Match segments differently.
  - Ensure that your team is informed about the segment statuses to prevent any misunderstandings.

Alternative Approach: Creating a Translation Memory (TM)
If you want more control over the segment statuses or prefer not to use Perfect Match, you can create a TM from your partially translated XLIFF files:

1. Create a TM from XLIFF Files:
   - Go to Translation Memories, and create a new file-based TM.
   - Import your XLIFF files into this TM by selecting Import and choosing the XLIFF files.
   - This will extract the translations from your XLIFF files into the TM.

2. Recreate the Project with the TM:
   - Create a new project with the original source files.
   - Add the newly created TM to your project settings.

3. Pre-Translate Using the TM:
   - Run the Pre-Translate Files batch task.
   - Set the appropriate options to leverage the TM matches (e.g., specify minimum match values, segment statuses).

4. Run DeepL Plugin on Remaining Segments:
   - Use the DeepL plugin to translate any segments that didn't receive matches from the TM.

Advantages of This Approach:

- Control Over Segment Statuses:
  - You can specify how segments translated from the TM are marked (e.g., as "Draft" or "Translated"), providing clarity for future reviewers.
 
- Avoiding Perfect Match Status:
  - Segments won't be marked as Perfect Matches, which might simplify certain workflows or QA processes.

- If maintaining the exact segment properties and statuses from the initial run is important, and you're comfortable with how Perfect Match affects your project, then your initial approach is appropriate.
 
- If you prefer more control over segment statuses or want to avoid the segments being marked as Perfect Matches, creating a TM from your XLIFF files might be the better option.

Aries S. Patrick Andrew Hartnett 

Moved to a more appropriate forum.

Thanks for the very helpful reply Patrick.