The problem you describe was the subject of a talk with title "Who's afraid of XFA?" video/slide deck.
The solution posted by Max Wyss (fill out and print the form using Acrobat Reader DC) is certainly possible if you have only a limited number of documents that can be processed semi-automatically. If you want a fully automated process, this typically isn't a viable solution. For instance, if the form needs to be filled out in the context of a web application and people need to get the PDF in their browser as a response to a request to your Linux server, you won't use your Macbook as a server, will you?
In that case, you need a library that injects the XML into the PDF (see How to fill out a pdf file programmatically?) and then "flatten" it to an ordinary PDF (How to flatten a XFA PDF Form using iTextSharp?). The problem with this approach is that you need a software library or tool to achieve this, and other than Adobe LiveCycle and iText's XFA Worker, I don't know of any tool that can flatten XFA to ordinary PDF.
Using a PostScript driver to print a dynamic XFA form won't work because other than Adobe and iText software, there is very little support for XFA in the software world (and XFA is being deprecated in PDF 2.0).