What do you mean by "I don't actually know the format of xml"? How will switching parsers (though, technically, XML::Twig is a front-end, not a parser itself) fix that?
XML Twig saves on memory by only reading in bits of the xml. To know what bits you are interested in you have to tell it.
As I don't know what tags will be used for a given document I can't do that so it would load the entire xml into memory which negates the point of Twig.