The main distinguishing feature of the hypothetical archiving system I was looking for is
the use of internal redundancy to recover from the effects of occasional storage or
transmission errors not detected/corrected by the underlying disk drive/filesystem/network
etc. In this way, each physical copy of the archive could be independently replicated and
transmitted reliably without having to fall back to a completely-redundant and perhaps
geographically remote copy in order to deal with these increasingly routine events.
I haven't had chance to look at it carefully, but this looks like it is addressing the
issue. They are applying RAID-like methods to allow recoverability from isolated errors
with an efficient degree of redundancy.
http://www.quickpar.org.uk/
http://parchive.sourceforge.net/
--Bill