That's a horrible comparison... The better comparison is the H100 and NH-D14 as cbrunnem mentioned. At least they're closer in price and performance.
Regardless, I agree with muffin here. If you want to go water, go all the way with custom. No need to fool around with closed loop stuff.
And temps depend on a lot of things (ambient temp, voltage, etc.), so we can't give you an exact answer on that, just our own experience (mine being that I have core temps from 59-63C in Prime 95 @ 4.5 on my 2500K with a 212 Evo in push/pull). But I've seen other people with lower AND higher temps, so there's no set range for each cooler.
well i do agree it's unfair to compare the two, the h100 will obviously be better but i want to see if the h100 will be overkill for me. I also don't have the time or money to go custom. Basicly I want to see where each coolers hit there limits.
BOINC (which runs SETI@Home, MilkyWay@Home, etc) are very similar for loading, but Intel Burn Test works better than anything I've found, including Prime. It's brutal for stability and load testing...I haven't found anything that loads a CPU harder.