~
Only post worth reading in this entire thread.
The economics of the fee attack are obvious. They do not need to spend that much money in order to get the fees to go up. If they flood the lower fee levels with transactions they are paying smaller fees if the transactions do go through at all. Also, due to their own attack their transactions spend a lot of time in the queue just clogging up the queue - less cost for the attacker. If they are really successful then most of their spam transactions will never get into a block and will eventually get dropped. The cost to the spammer of the transactions that get dropped is zero.
~
that is not entirely true.
in order to have an effect on what others pay as fee you DO need to pay higher fee or at least a fee on par with others.
the fact that bitcoinfees website shows majority at 101-120 is because many are using old or bad wallet clients that will suggest a fixed default fee of that amount. for example blockchain.info is always going to say 110-120 satoshi per byte.
if the attacker floods the "queue" with low fee transactions they are ALL simply going to be ignored.
for example if there are 100,000 transactions with 100 satoshi per byte fee, i am going to pay a higher fee of 110 s/b and have more priority than them.
in order for a spam attack to be successful and cause higher fees and delays it needs to pay higher fees than what "I" pay. and they do pay higher fees. i have seen from 300 to 600 s/b (there have been 24993 transaction in last 24 hours with 301+ fee)
and it is not just weird unknown entities spamming, it is miners who are filling (wasting) blocks so that they have less space for real transactions.
186704 tx belonging to Bitfury: 3QQB6AWxaga6wTs6Xwq8FYppgrGinGu15f