Need to revisit Tape model in the out years

[StevenCTimm]

There are 2 changes that need to be made
(at least.I think they still need to be made).

One--there's a transition where CERN tape turns off a few years after protodune which does not show them hanging
on to the protodune data they have.
Two--about the same time the Rest of tape goes to a split 50/50 between FNAL and "National"
and as part of that the amount of tape at FNAL actually goes down.. unless that's been fixed already.
Not realistic to ever see actual FNAL usage go down, and also there are other discussions where we've said that one-copy stuff will be only at FNAL.. so have to revise those assumptions. Every time existing model has been shown at Fermilab Jim Amundsen has laughed at it.