Good comments about the details of the protocol. But I'm wondering why nobody found anything to comment about it on higher level than the crypto? We all know(?)

Good comments about the details of the protocol. But I'm wondering why nobody found anything to comment about it on higher level than the crypto? We all know(?) that Tor and multihop data passing isn't efficient way to implement 'anonymity' for distributed file sharing. For that particular reason I was personally amazed that they did select Tor as example. Tor wastes a lot of bandwidth as well as allows easy traffic correllation attacks in the cases where that's generally feasible. I really loved Freenet and GNUnet designs, because those use really efficient caching, partitioning, routing compared to Tor. At least in theory anonymous downloads could be even faster than when using non-anonymous downloads, due to improved efficiency of the network resource utilization due to distribution and caching. When Tor is used as base, all these benefits are lost and in addition there will be huge bandwidth overhead causing about 600% slowdown. Does anyone agree with me? I was almost sure that someone would immediately comment this aspect, but as far as I can see, nobody has noticed these facts(?) yet.