At this time it is mostly correct.
But once caching is implemented there will be a lot of nodes supplying the chunk if it is popular.
It would end up being a very bad seeder as files drop off BT network and your Autonomi seeder would be claiming all these computers have the file while they have not for who knows how long.
While I am sure there will be attempts to do this, the issue is outdated seeds.
Also if Autonomi is a choke point for popular files then also will become a choke point for seeding, just not as quick. And when caching is fully implemented then the choke point disappears. Caching of chunks (immutable records) is a simple concept and would use similar methods to replication. The nodes used to help find the closest node to a chunk then also grabs a copy for caching and removes cached records according to some caching algorithm which determines when its not useful to hold it.
So not too hard to implement really, just not needed at this time and the development/testing effort is going towards getting the basics working efficiently.