On Latency,
If My S1 GUi says the machine is running at 187 but the pool says 124 then that is Latency?
Maybe. Probably just variance, though. The "gaps" between new blocks and new work, and when work actually reaches you will cause a drop in your effective overall hashrate (your miner will continue trying to solve or submit old work, but they'll be stale by the time they reach the mining pool) - but as long as you're on DSL or better, it shouldn't make any significant impact (unless you have an enormous mining farm where a .1% loss in effective overall hashrate translates to an >$500/mo loss, and then maybe you'll want to think about ways to reduce latency as much as reasonably possible).
For normal miners, I wouldn't be concerned about latency unless a ping to the server takes >100ms. A significant farm, maybe >50ms - and a mega-farm, maybe >15ms (they should probably be running their own pool). Fiber connection to a server within a couple hundred miles should have a ping around 5-20ms, cable will probably be 15-40ms, DSL probably 60-120ms (3G 80-250ms, 4G 40-140ms, satellite 80-500ms).
ETA: to add to W2W's earlier post - stratum is MUCH more gentle with latency than how it used to be. You won't need to be getting each piece of work from the server now and basically just keep going without constant communication until you receive news to switch (and hopefully, you've submitted all your completed work prior to getting news to switch) - so getting the news to switch ASAP and making sure your own work is submitted prior to that is still important, but bandwidth requirements and the overall effect of latency is much reduced.
