It sounds as though you had some bad luck there.
If the two databases are being merged based on a person's name, there are inherent challenges. For instance, If John Robert Smith goes by his middle name, he may appear on different lists as "John Robert Smith", "John R Smith", "Robert Smith", "Bob Smith", etc. If the match is based purely on name, those variations will result in duplicates.
If this is a match by name, there are basic techniques that can be used for preparing the names for matching. I will use those techniques. However, this price does not include manual examination of records to search for duplicates, unless the record total is < 2000.
It's likely that the incorrect dates can be fixed, and the missing data included.
If you have any questions, please feel free to ask.
Thanks for your attention!