I came across two papers written on this topic:
I am probably missing a lot of the nuance, and I could try to summarize here how I understand it, but it'd probably be wrong.
1) Are these systems feasible? Feasible in the sense that I could access or send information to a web application and not require more than a few seconds in latency.
2) Would these concepts work with existing infrastructure? Or does it require being entirely removed from the way internet traffic is handled now?
3) Do these systems have any sort of extra costs to them? Does it create a tremendous waste of bandwidth or processing power?
4) Are there any more resources on this topic I could look at?
5) Is moving away from HTTP a pipe dream? Does anyone have any real motivation to do so in order to have more privacy/anonymity online? Is there a better method for privacy online that doesn't rely on using proxies or concepts like Onion routing?
6) My impression is that these anonymous message delivery concepts are theoretically a more secure (in terms of hiding identity) concept than Onion routing. Onion routing seems like it would have tremendously more ways to be thwarted than something inherently anonymous. Is this wrong? Is Onion routing anonymity superior in some way?