I've been wondering if the baptism numbers are 185600 rather than 18560 and somehow a zero has been accidentally lopped off.
This seems unlikely to me because this number is derived from the difference in peak publishers year over year, not baptisms. For the number to be off in such a way the mistake wouldn't have been a missing zero it would've been under reporting peak publishers by 167040. That doesn't turn out to be such a tidy off-by-one type error.
Furthermore, looking at the numbers for 2014 and 2015 together it paints a clearer picture. There was a larger increase in peak pubs in 2014 (probably due to their big august push of jw.org flyers combined with anticipation that oct 2014 might see armageddon) than usual. Because the factors leading to the 2014 increase could not be sustained, the numbers corrected and dropped back down. Taken together the average increase for 2014 and 2015 is ~126k increase per year which is more in line (but still a good bit below, in most cases) previous years.