The way I read it, the individual forgers will choose a minimum fee they are willing to accept.  

The individual users will select a fee they are willing to pay.  As long as the fee of the user is above the fee of the forger, the transaction goes through.  The forger will get all of the fee offered up by the user, not just the minimum.  

The the users fee is too low, then the transaction will wait for the next block or for any block that will process it.  

I am not really sure if this is going to be implemented or not.  

Originally, I thought it was strange, but after thinking about it more, it is a really nice system.  After a while the free market will sort out an average known price for making sure a transaction goes through.

The problem with choosing a set price like 1 or .01 or .001 is it is so subjective and will constantly have to be changed as the price of NXT goes up and down.  

Static fees just don't represent a free market.  

How much have been made with mining?

Just reading your economic paper: https://docs.google.com/file/d/0BwAGADgnQcrtdXE5MkF5S05oaHM/edit?pli=1

Some questions:
-Do you want Nxt to be used as money (medium of exchange + store of value)? Because on page 4 I read "Deflation is not much better than inflation. 'Real' coins should be created on top of Nxt, and be issued in quantities that keep their value constant."

-Do you want to increase the amount of 1 billion Nxt? On page 4 it says "there is no limit for a specific type of monetary expansion".

I really like Nxt, but these economics discourage my enthusiasm

Oh wow.  I hadn't heard of this before, but yeah, perfect idea.  Let the free market decide on the fee like it should be.  I hope this is implemented as well.  I don't see any point in waiting though...

nxt should always be used as "the money" inside of the nxt ecosystem because the security of the network is atleast partially a function of the price of nxt.

the 1 billion amount can never be changed except with a hard fork. just like bitcoin.

I see no good reason to expect wealth concentration and a static community of holders. I mean its true that a large stake holder will tend to make larger transactions than a smaller stake holder meaning a smaller percentage of his wealth will be consumed in fees, but you need to also assume that larger stake holder will tend to redeploy a smaller percentage of their nxt worth than a smaller stake holder. Do we we have any data to support this? Granted wealthy people tend to consume a smaller percentage of their net worth but are they not, in general, more diversified in their investments than less wealthy people?

The point is that, maybe its true, im not saying that it isnt, but we simply lack the data necessary to come to a determination either way.