another thing I observe is when the difficulty changes seems to be some correlation to long block times, as if it takes the pool time to adjust.
There, ma friend, you have a point! My observation about that particular conspiracy is congruent with the spirit of what you say , my minor divergence emanating from the little detail of "
long block times", as the reverse can also be true, indeed it happened not long ago. Remember when we had those 10 blocks in a day? That was on the 11th Feb 2015 and that
lucky run began a mere 10 hours after the increase in network difficulty from 41272873894 to 44455415962.
So it seems pool luck can be influenced by how fast the algorithm dishing out (and co-ordinating) the work adjusts to a new network difficulty (if ever there is anything of the sort!)
PS. And if there is such a thing in existence as
optimisation of the algorithm dishing out and co-ordinating work, then please do tell since I'd very much like to implement it in my very own proxy for solo mining ...!!