I normally don't have to deal with any OCR documents, I'm sure that would be a pain if most of my documents were old books that were scanned in. I think 98% of all of the material that I have to read are written post 1990, so virtually all of them are available in some soft format or another, most often in PDF or some Office format, occasionally epub.

With regard to Greek, most of the greek symbols that I read are pretty well supported in PDF creation software like Acrobat or Latex compilers. Actually, I used to study Attic and Koine Greek over several years, and I'm sure that a lot of those texts are only available in OCR from googlebooks. But I think I can obtain most of the well known text these days in html from somewhere, and view it in a browser, so it's not such a big deal. And I think the mathematical greek symbols use a different set of encoding than that of Homer through Byzantine Greek texts, if I remember correctly. I don't know anything about modern Greek though, don't know if they still use the same accents and diacritics. Anyways, I think support of fonts should get better over time on any platform, as more localization options for an OS become available over time. I'm also hoping to be able to read some German texts out of the box, fingers crossed.