as you can see in the image the OCR results are not that good here. Is the text too small? When doing OCR, I activated English and German as the languages and high quality. Slightly bigger text on white background had better results but they were not good, too.
(To make the text visible, I moved the new layer to the bottom and removed text formation.)
I finally tested the new version 317.1 of PDF XChange. It is way better than before. The right picture still contains ~3 mistakes per line and the 1:1 sized picture of the "About" has really weird results.
Left example uses the settings: High quality Language only English.
Right example uses the settings: High quality Language only German.
When you use "Medium" the OCR tool relies more on dictionaries, and when you use High - it tries to recognize each letter on it's own - so indeed for normal text - Medium gives better results. For other unusual strings (e.g. license keys of some ID numbers - letter and number combinations) - High might be better.
