The current implementation (as far as I can see) leaves very little trace of the original instrument-file and / or notebook used to perform the simulation.
I believe it we will find it useful to have access to the "full data chain" and all metadata for future debugging.
(Don't care what the dataset / location is, but the .instr file and parameters used should be left in the file.)