Even if you take a small chunk of text, though, you will still find repeated letters at about the same distribution as the Obama birth certificate. For example, looking at just the first line of the first Korea document (1952-01-03a.pdf), the o's in "South Korea" are identical to the pixel, as are two of the three e's in "revealed."
I still think this is a common occurrence in PDF files of document scans, particularly those that have some enhancement to the text, be it OCR or something else. It's common enough for me to conclude that it is not a sure sign of tampering.
Those Korean pdfs and even the books would be items where a conventional OCR, or some similar variant, would be utilized. Most of the pdfs I found of your example title had been fully OCR converted. But comparing it to Obama’s COLB is apples and bowling balls.
It simply makes no sense after having gone through such absurd steps to insure the COLB arrives to the White House unsullied. They hand it over to some intern hack to scan in using whatever protocol applies to the daily mail. It’d be like me paying max dollars to ship my Lamborghini safely to my doorstep, so I can park it on MLK BL.
If you want to believe that something might/maybe/could happen, IF, this, that, or the other thing also happens, feel free. I for one, am done making excuses for Obama.
Ok. I’ve looked at that first Korean pdf. That entire top row has the same “o”. Then the rest of the doc is different.
IMO this is a clear indication of manipulation. But here’s what I think is happening. They’re using a captcha system. When a character or word is getting kicked, the operator is simply substituting it. FYI Google books (your example) uses a captcha system for their ebook pdfs.