I not too long ago had to "un-correct" a merge that someone had imposed on my own ancestor. The names of two (2) people were identical in the mind of the merger, although one had a middle initial and the other did not. The birth-dates were in fact identical, but stated only as a year with no month nor day listed. More importantly, there were two (2) tombstones in established cemeteries in different counties of South Carolina (United States of America), and those tombstones declared two (2) different dates of death. In spite of this, the impatient merger proceeded to merge these people in my "open" listing even after I pleaded for delay. In fact, these two (2) persons were first-cousins in families that had upward of a dozen children per generation. In this area in the late 18th century, given names frequently were recycled as honorifics (or possibly due to lack of imagination or willingness to go against local cultural norms). (A possibly peculiar local practice (that seems to mystify modern folk who live in other regions) occurred when a child was given a name, the child died, and the very next child of the same gender was given that identical name.) I urge less presumptuousness when reviewers encounter apparently appropriate merges. Some of us do not have the time to respond immediately to proposed merges, nor the time to correct errant revisions. In this particular case, if the person who proposed the merge had considered the obvious geographical and terminal date discrepancies -- or if hin had been willing to simply search for redundant names in the "tree" that was being reviewed, the problem could (indeed, "should") have been avoided.