Customer specific tests depend on whether the document was scanned or produced from a digital source and how it is intended to be used. For scanned documents the following checks could make sense:
- Resolution of the scanned images
- Compression algorithms
- Manufacturer of scanner (stored as the producer property in the document's metadata)
- Presence of OCR text
For digital born documents the following information could be helpful:
- Names of the used fonts
- Font embedding information (if not PDF/A)
- Creator and producer application name
And, in general:
- Page format (A4, Letter, etc.)
- PDF minimum and maximum version (e.g. from 1.4 to 1.7)
- Presence of specific features such as embedded files, transparency, patterns, shadings, color spaces etc.
I'd like to learn more about your specific requirements. Please let me know them and post a comment!