This is an interesting one. The PDF does not have appearances (the representation of the field values on the page) for the fields. The values are there, just not what they should look like. Based on the PDF specification, the tool that populates the data should generate these appearances but isn't. Unfortunately, Embed API isn't generating them on file open either. While technically this isn't a bug because Embed API is following the behavior defined in the spec, I don't think it's a great end-user experience so I've submitted the issue to engineering.
I had to change my flow. What I did was, I built a template in Microsoft Word, filled that template with Python, and finally convert that docx. to PDF.
In my first post here, my flow was, that my template was a fillable PDF created in Adobe Acrobat Pro, and I filled that template with Python and save it and unfortunately show up in my web app as I explained.