Backlog overwork is proportional to unused code?

A common problem I see in my travels is backlogs where items in the far future are excessively detailed. A subtle dynamic related to this overwork waste is that the more detail a developer knows about future stories the more likely they are to add extra "flexibility" for them into their current story development. And of course this contributes to the amount of unused code in the software (which is often quoted at averaging 45%). I therefore propose a hypothsis:

Backlog overwork is proportional to unused code.