Way back in 2003, I thought I had "figured out" an 1860 census entry with a few irregular entries. I even had a list of reasons why my conclusion was correct.
Flash forward to 2012. In attempting to "redo" the research, I reached a different conclusion about the 1860 census entry--one that meant I had more work to do.
Genealogical conclusions are always subject to new information, new procedures, and the potential that a misinterpretation was made along the way. Don't be afraid to revise.