Sorry, I'm not a native English speaker, so I don't get what your impression is from the words I've wrote. But as far as I understand, you seem to acknowledge that there are many uses for EPUB files. So I can't imagine why it shouldn't be a good concept to validate EPUB input files before further processing in order to make sure that the code doesn't run into unexpected constructs or even non-well-formed XML. It's actually quite common for all kinds of XML applications with standardized or custom XML Schema format specifications, because XML validation can be done off-the-shelf with programming libraries and it clearly points to the various errors in a XML input file, which need to be fixed anyway in order to produce quality output that can be used as input for other applications in the future.
Edit: I really want to encourage everybody to get familiar with the process behind EPUB standardization, see “
Introducing ePUB3” by Bill McCoy and “
EPUB3: Not Your Father's EPUB” by Bill Kasdorf if you haven't already.