I was able to calculate a little more data about whoever has been cherry picking the high-value blocks. I was a little concerned that the times I showed might be misleading. They aren't the times the cherry picker needed to solve the blocks. They are just the times between the blocks he solved. Time that he wasn't even trying to solve ECC blocks would be in there. Also, I caught a minor mistake. It turns out that block explorer sometimes returns blocks out of order. Assuming that the blocks are numbered in order, so block 12345 should be after 12344, it turns out that sometimes he solved blocks one second earlier than he solved the previous block, according to the official block data. Don't ask me how that's possible! But anyway, the minimum time in the table above should have been -1 seconds, not zero. If you include the blocks that he didn't solve, then it drops to -4 seconds. This doesn't look as bad as the KGW timewarp exploit, at least not as far as I know, so I wouldn't jump to that conclusion.

Anyway, here is the new data:

         ECC /    Time Since   Time Since   Block Age
          Block    His Block    Any Block    His Block
Average   239000    0:02:39      0:00:20      0.0002
Median    248109    0:00:44      0:00:11      0.0001
Min       100332   -0:00:01     -0:00:04     -0.0001
Max       299862    0:25:03      0:04:29      0.0032


Note that we don't know when he starts trying to solve a block. One time, he solved a block 4.5 minutes after the previous block was solved, but when did he start working on that one? Perhaps just a few seconds before he solved it.

Note that the median time per block, if he solved it, is only 11 seconds!!! That's why it's tough to compete with him. You should expect 45 seconds per block, but you need to find a solution in a fraction of that if he's around. If the block has a high reward, he will likely be around. Not all of the time, though. Apparently for 25 minutes he found something else to do, no matter how high the ECC reward was.