This may or may not be desirable, but it seems unlike ArticleInfo, Top Edits does not discount reverted edits when computing the total added text.
See the discrepancy with User:Ozzie10aaaa: On ArticleInfo they are reported as having added 31,252 bytes, whereas on Top Edits they are reported as having added 35,940 bytes.
If we wanted to make them consistent, the Top Edits tool might be slowed down a bit, but it might be worth it. Even better, we should show both statistics, and include the number of edits that were reverted, and in the table highlight these edits. That would be pretty nifty, I think, and would clear up the discrepancy.