I tend to think that any disease that is diagnosed without a well understood mechanism and a physical marker is very commonly misdiagnosed. We may find out depression and CFS/ME are two flavors of the same thing, or we may one day prove they are distinct and that many diagnosed with depression have CFS/ME.

I think this applies to all the diseases that are diagnosed as a bucket of symptoms. For example headaches...someday when we understand the mechanisms behind migraines, cluster headaches, tension headaches, etc... we'll probably find that we drew the buckets in the wrong spot.