was destined to not scale from the day it was thought of. There have been years and years of research into making it so PIR scales, everybody gets everything is least scalable possible. BitMessage has problems from the bottom to the top, and some of the problems it used to have were so bad that it obviously wasn't made by people who have a clue what they are doing. Even now that they don't encrypt the entire message with RSA they still have tons of traffic analysis issues etc. Pretty much the problem area they are working in has had truck loads of research into it for the past fifty or so years and they obviously didn't read any of it. I learned the same lesson some years ago, don't try to invent your own secure communications system if you are not an expert, and you are probably not an expert if you don't have a Ph.D or at least masters degree in a related science, or if you have not spent 5-8 years researching on your own, while having a solid understanding of advanced math as well. There already are hundreds if not thousands of whitepapers out there from people who are experts on such things, if you want to make a system read their whitepapers and spend your efforts implementing them (which is hard enough in itself) and thinking of ways to make their systems work together (because a complete system for secure anonymous communications is going to involve several different parts, and each part will have its own whitepaper). The less that you need to invent yourself the better. They took Bitcoin and tried to make it a message system, so they didn't totally reinvent the wheel. But they thought that everybody gets everything was a good idea for messaging systems to maintain anonymity, and theoretically it is, but having that thought without knowing what Private Information Retrieval is will lead to ignoring the fact that there are decades of research into having the same anonymity properties while using orders of magnitude less bandwidth.