I suspect these decisions can be complicated, particularly when you have an on-going development cycle & production line that you don't have to support, and are facing the pending removal of support for your current type.
It may sound odd but options such as buying a smaller than usual number off the shelf, flogging them to death for 10 years, then scrapping them, could work out attractive from both a capability and cost point of view compared to a more traditional 25 year buy and all that entails. If the US are paying for the bulk/all the 64D to 64E (and potentially future 64F etc.) development costs, it may make sense. Some DLoDs may complicate issues of course.
Airlines have a habit of flogging ac as they approach e.g. a big D check - not quite the same, but principle is similar.