FWIW, I suppose one way to identify and correct errors is to compare the text from wikitext with the text from
scriptures.byu.edu.
I assume
scriptures.byu.edu could also have errors in it also. But I'll bet that BYU's HBLL paid students to correct the OCR job at some point (I know I'm making assumptions on that, but it seems like the way they would approach this).
In any case, even if they are both just two different OCR jobs without correction, comparing the two will catch errors that other methods (including proofreading) might miss.