Or better yet, a comparison pane/window that shows just text?
I am concerned with historical scans and public-domain uploads and users being able to copy and paste and possibly translate. Viewing the document/image, I can highlight and type, but I know there are many errors I just cannot see.
It's a pretty big deal and seems like it should be an easy feature for this advanced app to carry.
It took me a few moments to figure out what your wish was related to, and I think I now understand that a word "looks" correct, but does, in fact, have incorrect OCR results.
I am aware of two options: The Acrobat way to correct OCR results is after (or any time later) the OCR process is done, go back to the Scan & OCR tool, and in the middle of the top region, you'll see a dropdown for Recognize Text. If you look at the bottom of that dropdown, you'll see an option to Correct Recognised Text. Select that.
Once that opens, you'll see this: The text in question is highlighted and on the top (screenshot showing this), you'll see a replication of the word(s) in question and a window to the right with the option to Accept the words in question or Accept any correction you make.
Depending on the nature of the document and the number of questionable words found, this can be quick or deadly slow. One of the features it needs is the ability to auto-correct any similar mistakes. But alas, this option is not here.
One of the things I have done is to either copy the text, or export it into Word. Then, Word can display the red lines under the misspelled words. You can then look to see how many words you'll have to deal with. You can also try to fix the Text in the page in Edit mode, but I've found this to be somewhat problematic by the way that Acrobat displays the font depending on how you set the output in the OCR process.
I wish I could tell you more encouraging news but correcting miss-OCRed text in Acrobat is not always easy (although it is better than it used to be, if that is any consolation).