Bytemaster, appreciated your hard work and briliant idea. I think it's a really good thing that we weight all the proposals based on objective oriented instead of religiosely for the ideal concept of pure decentralization.
Although it's maybe a problem for marketing since the competitors will attach this concept w/ centralization and single failure point(whatever it's true or not), I believe if we really find a way to resolve all the business logic and technique issues or make it the best balanced solution obviously, we could educate people to support and join it gradually.
Till now, the only issue really bother me is the one I quote below - DDoS. I know as you mentioned the notary node(or processor node..) would have redunant all over the world(not just likely, it should be a requirement for "the notary"), but I still have 2 concern:
1. compare to the whole alive nodes on the network, the quantity of the reduntant nodes of "the notary" is
lower by ord of magnitude. To DDoS ten or one hundred nodes takes a lot of resource and much more
difficult than attack single node but it become feasible to do.
2. It's true that to attack all the reduntant nodes simultaneously is very difficult, but how about attack them
one by one? The attacker DDoS the current notary node first, and after the notary aware of the attack and
start to use the reduntant node, the attacker also start to DDoS the new one. That will impact the trade
very seriously and the attack could leverage it to affect the market price then make money from it.
This 2 concern will become more realistic when the system grow up to carry huge wealth and the expectation of this risk will suppress the value of the system seriously..
- One trustee one point of failure, the DDOS attack will be hard to prevent
DDOS is not really possible because the network would still be broadcast based and the trustee would likely have multiple redundant nodes all over the world ready to sign blocks if there were an issue.