The fonts used in a Topaz book are often not that good for reading. So conversion is usually my best option. However, I find that the OCRed text is usually terrible and requires a LOT of tweaking. You will find the usual OCR errors such as T being rendered as "I" or 'I', G being rendered as C and vice versa, E being rendered as F and vice versa, et cetera. It is usually worse if the original scanned text was from a serif font.

At any rate I usually just convert to HTMLZ and then open the html file and run search and replace on those typical OCR errors and any others I can find. I rarely buy Topaz books, otherwise I would try to find an app that will automatically do search and replaces on common OCR errors.