Why would I want this?
I can think of no value in a mass of ocr'd screen shots that is remotely worth the cpu and disk cost of generating them or the security risk of them merely existing anywhere, even if the code to do it was gpl.
And the real problem is not me anyway. I use linux. But not one of my relatives I care about does. They are all going to have these terrible things. I can directly take care of a few of them but not most.
Not to mention just everyone else everywhere.