Although  this plan  sounds reasonable  on  first hearing  (given the  present
"environmental" conditions) it's not a workable solution.
Optimistic view:
Let's  assume LISTSERV/LISTEARN/LISTwhatever  turns into  something like  MTAs
(mail transfer agents)  - so let's call them LDAs  (list distribution agents).
Obviously there has  to be some standard (RFC4000-RFC4999)  which is supported
by all  LDAs. This may  seem unnecessary  work but on  the other hand  it will
enable non-VM  sites to implement LDAs  and connect them as  peers to existing
VM-LDAs. Sure this will take some time  and meanwhile the net will be split in
to parts in terms of LDAs.
Less optimistic view:
A similar  scenario but only  a minimum standard can  be worked out.  EARN and
BITNET/NetNorth agree on  the use of special RFC822 tags  and everything works
more or less like the ARPA redistribution lists (with similar restrictions).
Pessimistic view:
NJE networks  are already outdated. So  why not start right  now? With mailers
already supporting  various flavors  of routing  and TCP/IP  and all  kinds of
gateways the RSCS-based network is an anachronism.
*-*
With the  proposed plan  a (partial)  functional split  of the  networks seems
unavoidable - a situation I dislike. I also don't know if BITNET/NetNorth want
to make themselves headaches over something  which is not a real problem. I.e.
I don't know  if EARN made an  official request to talk  about the efficiency,
functionality and usefulness of LISTSERV (or an equivalent tool).
 
Christian
P.S.: this  is  not  a  final  statement but  meant  to  further stir  up  the
      discussion.