I am trying to set up a site that will be storing a large amount of images (roughly 50,000 to start with). The strategy for this was to use Storage API and S3 storage to deal with the costs of hosting, and Services API to provide an interface for a offline converter/uploader to get the images to the site.
This is working pretty good. If I upload an image manually by creating a node and filling in the image field, it gets sent to Storage API and good things happen. I can also upload files and create nodes from the offline uploader via the XMLRPC Services API.
The problem is that the files uploaded with the offline uploader don't get handled by Storage API. Currently the script uses the Services API method 'file.create' to create and object, then the details returned are used to fill in the image field on the node that is created next. I expect there is a better way to do this, but I am not sure.