this is only bad in my opinion because it proves one thing and shows a couple of more bad things.
first it proves once again that there was in fact a spam attack causing this massive backlog.
and it then shows that how mining pools are ignoring the transactions with low fees and only mine those that have high enough fee for them. because if you look at the number of unconfirmed transactions during this time it never reached zero or near zero. it has always been above 5K but when there is no more high fee transactions they MUST include all the low fee and zero fee transactions. and apparently they choose not to!
for example right now there is 8800 unconfirmed tx but last block picked up 1349 or 695 kB of them.
p.s. there is no migrating when there is nowhere to migrate to
Never had a single problem by selecting the recommended fee by the Bitcoin Core full node program. The backlog was indeed due the massive spam periods by Roger Ver camp.
Now that they have BCash to worry about they are too busy to keep spamming. We may have had increased sizes recently due people moving a ton of BTC into exchanges and other wallets in order to dump the BCC (BCash) for more BTC.
Let's get 2MB blocksize in the future, not in 2 damn months. Please NYC participants, THINK again if you still think it's a good idea.
Let's allow segwit to run for 1 year, see what happens, then consider a proper hardfork.
Let's not mess with a nuclear reactor in a stupid way.