I remember a massive clusterfuck happened with NXT, and it was due the PoS model being exploited, namely the "nothing-at-stake" problem which was never solved as far as I know.
Can you give more details? I don't remember that event, and I was involved in the NXT community for a while. The most close thing was a BTER hack, but that was a simple "exchange hack" and had nothing to do with the protocol. That hack
could have been exploited for a PoS attack. However, the hacker would have had to buy about one time more the hacked amount (afaik the hacked amount was a little less than 10% of the supply, and he would have needed 15-20%).
AFAIK the nothing-at-stake problem was used only in an attack to a very small coin (
APEX, attacked successfully by cynicSOB).
Overall the top people in the field have concluded that they wouldn't trust their coins in a PoS coin. The distribution of the initial supply is problematic, and the several exploits in the long term don't make it strong enough to hold anything long term.
I agree with respect to the distribution of an initial supply via an ICO. However, the initial supply can be distributed via PoW like it occurs in Peercoin. N@S is an ongoing debate, there are multiple approaches like Cardano, Casper, or Snow White to make PoS more robust against it - at the end, the big question is if a N@S attack is easier or harder than a standard 51% PoW attack. Both attacks are very likely so hard and unpractical that they will never have success in large cryptocurrencies.