For every reproduction there is a loss of information.
This is not a problem in museum studies but a fact of the reality we live in.
 A hypothetical Matisse that is copied stroke for stroke,color for color,
(and shame on you who think a photograph of a painting is as good as the
original since most paintings have something called texture), will still have
missing information such as the chemical composition of the materials used at
the time: Lead based Green, Cadium red  soon to be outlawed etc. Also you
can't carbon date a replica of an Anazazi pot for evidence on the timeline.
    If you chose to argue that the general public doesn't need access to this
sort of information and can make do with copies of the original than I
suspect you of having a limited view of your audience and perhaps should
allow yourself to be replaced by a virtual reality knowbot model of a curator
which would be cheaper and given your diminished standards, acceptable.
This was written by a facsimile