1. Can be also because of poor pool performance. If you try to manually connect to that pool and not through bitHopper, do your stales get less?
2. Is something I never experienced yet. Currently I'm mining away at ozco.in with:
[17:46:51] mining.mainframe.nl :3664244
[17:46:51] mineco :474659
[17:46:51] bitcoin.lc :5291446
[17:46:51] mtred :1915681
[17:46:51] triplemining :1046166
[17:46:51] bitclockers efficiency: xxx
[17:46:51] bitclockers :1404907
[17:46:51] bitp.it efficiency: xxx
[17:46:51] pool.bitp.it :1206567
[17:46:51] btcguild :600806
[17:46:52] btcguild efficiency: xxx
[17:46:52] eclipsemc :3799559
[17:46:52] ozcoin :358315
and it doesn't seem to have jumped anywhere. Do you use the latest version from github?
If you trade your dollars you also take a risk, i am fine with that.
I am upset about the fact that i am forced to either gain the system like you guys or switch to another pool with hopping measures in place.
If you find a way to increase the speed you could use it yourself and have little impact at all, sell it (again, little impact) or release it for free which wouldnt hurt anybody - word spreads fast and people would upgrade to your miner.
You would (if all other adopt this) increase the hash rate worldwide by 3%, thus you'd earn less than if you kept the tweak to yourself.
Pool hopping also has a risk of earning less in the short run (on long rounds you loose), pools switching payout systems, pool operators plainly stealing from you (which is REALLY easy by the way, look it up in the pools forum) etc.
If users are too stupid to be able to get a system like PPLNS (which is 100% like proportional, only that the definition of "round" doesn't start when a new block is found) or SMPPS (which is in the longer run just like PPS) then they also shouldn't complain if someone is smart enough to not mine full time in their pool if it pays off less and less over time.