Response by Nxt CoreDev Riker:
Wtih regards to the hackernoon article:
1. "Unfortunately, predictability facilitates DDoS attacks against the next minter. Also, we have to avoid that nodes can increase their chance of minting by manipulating the variables. That’s why we are not going to use this inequality as such but a more sophisticated scheme." - the NXT protocol protects the addresses of forgers (and now also bundlers for Ardor) so that DDOS attack is difficult to mount, there are various ways we use to protect against manipulating the forging variables. From requiring that stake takes 1440 blocks to mature, to not including the transaction hashes in the generation signature used to determine the pseudo random forging time. This issue has been heavily debated in the past and no real risk was identified
2. "That’s why many PoS cryptocurrencies avoid a long range attack of this kind by preventing long reorganizations. For example, in NXT a user may not accept the alternative blockchain if it differs from the existing blockchain in more than the last 720 blocks. This restriction, however, does not solve the problem for new users. When a newcomer connects to the network, he sees multiple blockchains with no prior knowledge of their authenticity."
- this long range attack has also been debated in the past, to start with you need to first obtain the passphrases of the 73 initial NXT stake holders. I think this by itself makes such attack impractical
3. Nothing At Stake issue - also was heavily debated in the past (not an expert on this). My understanding is that to actually profit from such attack on NXT you need to have a large stake, and if you have a large stake you don't have incentive to attack the network. But say you like to attack the network not for profit, still the damage you can cause by creating multiple forks is minimal unless you poses a huge stake.
If you have any other questions, please join Slack : nxtchat.herokuapp.com