[...]but implying the pool could pay them sounds the same as saying the pool fees are high.
Not really, and pool does not pay them, it gets them back with the extra space on the next block. It's basically like this:
(To make things simple, lets say tx is 10kB, and the lowest tx fee of the last block per kB is 1mBTC, thus tx fee is 10mBTC)
- Pool pays for tx fees, i.e. 10mBTC
- On the next block, there will be 10kB more space for transactions. Thus, the block will be worth ~10mBTC more, that will go to the pool.
Pool fee stays the same, rewards for miners stays the same... but earlier!
Note: for this to work the tx for the payments needs to pay the same price per kB as the lowest paying 10kB of the last block (since it will be more or less the same as the next one, and its data already available) This amount is not gonna be 0.
If the reward was (magically) 10mBTC more, then the miners get 9.91mBTC of that ...
Me (the pool) getting 0.09mBTC more per block doesn't pay the fees ... which currently would need about 17.5mBTC per payout
... so I have to increase the fees, or running the pool suddenly costs me a whole lot more ... about 0.7BTC a month to add txn fees at the moment that supposedly would confirm in 25 blocks ... then when there's spam, that increases the cost per 1K to confirm, I'd have to increase the fees even more ...
Of course if any of our blocks contain any zero fee transactions (other than the payout) then removing the payout would not increase the reward at all in those blocks ...
... as I've already pointed out ... removing the payout will not magically add the same amount of mBTC in fees coz the end of the block is normally low or zero fee transactions.
Or we continue to do it as I do ... zero fee txn so no one pays any txn fees to get their (non-dust) rewards.
Since that's WAY better than any other pool, I like the idea also.
Even Slush decided to add that on his pool - I guess he liked my idea
... though he charges more than twice the pool fees I do.