So you are usually better of reading the code anyway because you cannot trust that some dev updating code updated javadoc as well.
When something goes wrong I usually have to do is to dig into GIT history and see when and what changes were connected.
The most reliable Javadoc is @author, git tells you the author, @author tells you who the code was copy/pasted from.
Names do become misleading. Of course making callstack deep enough, will hide that.
Code reviews have same power to rectify both: naming and comment issues.
Great point.. except for when you're dealing with a lead and/or reviewer that refuses to approve comments because "we write self-documenting code" yikes