Just to say, I had this issue with slow slow DOI saving. The query to look at existing DOIs and make sure the new one was unique was taking something like a few hours. After optimizing database tables, it was a few seconds. This was on a journal with very few DOIs, and on an install of OJS with about 4,000 DOIs assigned in the install (but not on the same journal I was working with). So, a possible alternative if the query is slow is to first look at optimizing database tables, and then see if it’s still slow before commenting out the lines that make sure that the DOI is unique.
I was in OJS 3.1.1-4 (so apples and oranges in many ways to 2.4.8), but still the checking that the DOI is unique and the concept of optimizing database tables may be similar.