I'm also interested in looking at the ATContentTypes migration machinery for walking through a site's content (for replication of content) and also the Zope3 based TreeWalker (see the tutorial on Plone.org) for the same purpose.

But, first steps. Add the script listed at the URL above, call it rpc_add and put it in the root of the plone site. , and then try using xmlrpclib ffrom Python on the command line. It's so easy!

>>> import xmlrpclib
>>> url = 'http://user:pass@server:port/path/to/plone/root'
>>> server = xmlrpclib.ServerProxy(url)
>>> server.rpc_add('/path/to/container','id','Title',\
... 'Body','Type','text/html','visible','Description')
'http://url/of/new/object'

How easy was that? The next step is to get it working from a python script on another Zope isntance.