I'm impressed by the amount of progress the team has made in the past few months, KMD is definitely a pioneer of its own kind. The only issue I have are the notary nodes - I always have an issue whenever voting is involved in any platform. You are signifying trust at that point (which is the completely opposite of what the basis of bitcoin was meant to serve, a trust-less service), and if there were to be a group large enough it would essentially become somewhat centralized. We can see a clear example of this by looking at Lisk.
Lisk has let this issue run rampant and now they have to deal with hiring a full "science" team to amend the issue, this seems like a greater feat in itself (rather than the javascript protocol).
None the less it's not like centralization really matters in the cryptocurrency market. Even BTC has ended up being severely centralized and people still use it. It seems in the end people tend to care more of the application or use. If that ends up being the case KMD is in my opinion a sleeper.
I think you might be assuming that notaries have some exclusive power to create blocks, which is what I think LISK and other dPoS systems have.
In KMD, the notaries can notarize, but all notarizations are validated by the other nodes, so there is no power to create false notarizations. There is no monopoly on creating blocks that the notaries have. In fact about one third of blocks are created by GPU miners.
So the notaries are stake voted into place and their notarizations are NOT trusted, but all are verified. Not so sure that this is any where near the centralizing issue that you seem to be assuming it is. True, notaries have the power to not notarize.
Thanks for the clarification. That makes more sense, how ever i question the integrity of notary nodes in a longterm situation, Since voting is involved. As you said, I under stand that they can't falsify blocks, but the majority of new kmd will be held by notary nodes, can't that lead to manipulation when it comes to voting (self voting and bribery).