I'd like to know about the technology on the back end. Judging by the output, it looks like someone just did a "Print to PDF" on OS X and then ran it through Acrobat Pro for the OCR.
Given that it renders my site completely atrociously for some reason I've not bothered fully investigating, whereas the PDFs generated by Safari on OS X look just fine (just checked), I can pretty well promise that they're doing something else.
No need for the OCR. It’s already a PDF with text, no image.