Some sync is that way, but there's also the (etymologically correct) meaning, about synchronizing things -- keeping two filesystems that are "the same" actually the same, without reverting updates made on either side. It's kind of the same issue as with version control systems for software development, and it's conceivable that (with homogeneous formats) you could use git as a sync engine; something like unison is more typical.

To know what type of sync is in order, it's not enough to know that the same format is used on each end. You also have to know what sort of changes are likely; if you're using a directory tree structure, a file-level sync that ensures the latest version of each file is kept probably suffices. If you're using databases stored in one file, you've got to deal with simultaneous changes in both copies, and go for something capable of record-level granularity.