Any time that you are manipulating or transforming XML, you should make yourself thoroughly familiar with two technologies:   XPath (a sort of query-language for XML); and XSLT (a way of transforming one XML document into another).   You should also understand XML schemas and the concept of validating a file against that schema.   The on-line resources concerning these topics are legion.

What’s the “win?”   Answer:   you might well be able to avoid most-if-not-all programming.   I think that this is one reason why Microsoft et al has moved to an external file-format that consists of (possibly compressed) XML content built against a formally-defined schema.   You might find that you can extract data from an arbitrarily-large XML document (or set of documents), and produce a sophisticated output such as a spreadsheet from it, without programming in any language at all.

Also do not overlook the fact that Microsoft Excel, in its current incarnations, knows how to use both an XML file and an XPath/XSLT as an input data-source ... especially when an XML schema is referenced and the file validates successfully against that schema.   You might well find that the product alone, perhaps with proper optional plug-ins installed, can deal with this sort of requirement with no external assistance from Perl or any other language (other than, maybe, the Visual Basic for Applications language that is embedded into every Office product.)   Fond as I am of Perl, let it be said that Microsoft Corporation has not been asleep at the switch!