^na it's real i have one i've been waiting on over 6hours ...zzz
Of course the blockage is real.
Over the past several months, I have had several transactions that took 10 hours or more to go through, and I had a few transactions that took more than 30 hours.
Each of those transactions had more than sufficient fees; however, they were done while some kind of a spam attack was taking place, and therefore transactions were delayed.
Yes, there are some kinds of expectations that blocks are going to be full (because of organic growth), but these major spikes are not blockages due to organic growth, but instead they are purposeful spam attacks, and that should be obvious from the level of their spikes.
Whether and how to deal with it, and how big of a problem they are may be another story.
I understand that sometimes folks are going to become concerned that their transactions are not going through, and in that regard, they may be prejudiced by such delays.. .especially, if they were counting on the transaction going through quickly (like within an hour or two)... One thing that I have attempted to do is to just prepare myself for the possibility that any transaction could be delayed by a considerable amount of time because of these kinds of blockages.
One time (about a month and a half ago or so), I had a guy selling me bitcoins during one of these attacks, and he was a regular face to face and trusted person (more or less trusted), and the next day when he came back to do the second transaction, I told him that the first one had not gone through yet, and since I trusted him, I did not mind doing another one, but I was going to charge him 1.5% more.. on the transaction. so it was up to him whether he wanted to do the second transaction. He said that he preferred to do the transaction because he needed the money, and we did it. The funny thing about that one was that the attack seemed to continue to be going on, and within about 4 minutes, the second transaction went through, but the first one was still pending for nearly 24 hours (at that time). Both transactions had more than sufficient fees, but the second one went through a lot quicker due to mere coincidence, it seems to me (the first one ultimately went through at about 31 hours).
No, I'm not trolling. Some simple math:
Say the network capacity is 15000 transactions per hour before the difficulty increase. Now the difficulty jump reduces capacity by 17%. So now capacity is 2550 transactions per hour less. Now assuming transaction demand remains the same, in 12 hours you'll amass 30k transactions above capacity which go into the mempool.
So that alone explains more than half of the effect. No need to postulate a "spam attack". Any slight (10%) increase in demand at the same time would account for another 18k transactions making that 48k transactions, almost your measure 55k.
I doubt that difficulty has anything directly to do with fewer transactions going through. The more important metrics seem to be how many blocks are being processed per hour and whether those blocks are full or not. I still stick by my assertion that such a sudden spike seems most logically to be a spam attack.