Cryptonite is highly experimental software and the concepts underpinning it lack extensive testing and analysis. Cryptonite is a fork of the Bitcoin core but the code has been extensively modified and expanded upon. It may look quite similar to Bitcoin on the surface but things work very differently under the hood. As such, you should use Cryptonite with caution and diligence until the code base has been properly reviewed and the robustness of the network has been established. We provide no guarantees or warranty, but in the event where something does go wrong with the network we will do our best to recover the situation and maintain all address balances so that losses are minimized.

If what they say is indeed true then they are in big trouble, first of all they are using an account tree + blockchain which I would guess only inflate the blockchain size in the beginning and only make it smaller as more dust transactions are pruned from the blockchain (effectively breaking it) not the mention all the huge security problems that are bound to occur at one point or another.

Our implementation of the reduced blockchain uses an account tree, a mini-blockchain (which isn't even a blockchain) and a proof chain, the proof chain is the only infinite part of the blockchain but it would  barely increase in size over years of use even at frequent blocks. Not only is our solution, as secure as the original blockchain, decentrallized, doesn't expose additional attack vectors and compresses the blockchain to be smaller than any other blockchain before by orders of magnitute it will be delivered tested and stable.

I can't say anything for sure since I haven't seen their whitepaper but it will certainly be interesting to see how it goes forward.

In any case let the kids be kids and let the scientists do their research in peace.