And is it just me, or does it seem like people keep remarking on the coincidental "pathological inability to find blocks" during each of the stress tests? (this is from memory, could be wrong)

Might it have something to do with how mining pools work, perhaps a bug or inefficiency that causes them to throw away or waste hashing power if they are busy collecting too many transactions, lowering the overall effective hashrate significantly? (And conversely the 0 tx blocks are using their real full hashrate??)

/no-idea-about-this-technical-aspect