Bitcoin and Nxt both have finite supplies and when most of the btc is mined all those miners will have to depend upon transaction fees for profit just like in NXT forging. The difference between the two is that with bitcoin their are high maintenance(electricity and equip) costs; profits from transaction fees are linear and competitive with new entrants adopting new technology everyday. With NXT you can effortlessly scoop up most the transaction fees with no competition further securing your dominance.
Not really. The rich don't get richer in Nxt. If you own 10% of the coins and someone else owns 0.1%, after both forging for six months you will own 10% and they will own 0.1%. It makes no difference. You only lose if you don't bother to forge at all. Now we have leased forging, there's no reason not to forge.
Lack of hardware costs mean a lower barrier to entry. In both systems you need money to get started. In Bitcoin you spend that money on hardware. In Nxt you spend it on the currency itself. (Or, get a job and asked to be paid in NXT, and raise a stake that way.)
There has been plenty of forks with PoS. None of them have really solved the issue of the History attack( not found in Bitcoin) and the solutions thus far all depend upon using centralized checks and checkpoints which somewhat defeat the whole purpose of a decentralized currencies key advantages.
Nxt doesn't use centralised checkpoints. It solve the history attack by not allowing block-chain reorganisations to reach back more than 720 blocks. This is a decentralised solution.
It seems to work. It is the coin with the third biggest market capitalisation so there is plenty of incentive to attack it, yet there has been no successful History attack.
Perhaps there is a solution to this security dilemma, but I haven't heard of any valid hypothetical whitepapers proposing one as of yet.
Lack of a decent whitepaper is one of Nxt's biggest PR problems, in my view. There is so much ignorance about how it works, and people confusing it with other PoS schemes and assuming it has features or weaknesses that don't apply. (I suffer from this myself; I don't know it nearly so well as I do Bitcoin.)
However, the current algorithm is published in source-code form. It's not secret. And there is a project to produce a new whitepaper.
A PoS currency would be far more interesting if there was a way to airdrop all coins evenly without investors and forging of coins rewarded tx fees a flat amount across every node regardless of the amount of coins they had.
If nodes got paid a flat amount, people would create millions of nodes, each forging with a single satoshi.
Airdropping the initial distribution without investors is possible for a new coin, eg with a "spin-off" that bases the initial distribution on Bitcoin balances at some chosen block. That's the only realistic way a Bitcoin replacement can happen.