More important yes, I wouldn't say much more important.
Well, that all depends on what you're comparing. If you're comparing two identical architectures, like DDR4 vs DDR4, or DDR3 vs DDR3, then it's no more important at all. A 100% increase in bandwidth gives you a 100 decrease in latency. A 50 increase in bandwidth gives you a 50% decrease in latency, etc.
He and I weren't comparing DDR4 vs DDR4, though. We were comparing DDR4 vs DDR3, and that's when latency starts to not count for much, compared to bandwidth. Let me illustrate for you. DDR4 2,133 cl 15= 14.064 ns of latency. DDR3 1,866 cl 9= 9.646 ns latency. According to the confused guy with whom I had been conversing, the 45.8% latency advantage that the DDR3 has over the DDR4 is a
huge performance advantage!!!111oneoneone
So, would you like to guess much this 45.8% DDR3 latency advantage gains the 6700k? Zero percent (or extremely close to it), compared to the 2,133 DDR4:
As you can see, a +45.8% latency advantage is more or less identical to a 14.3% bandwidth advantage, when comparing DDR3 to DDR4.
DDR4-2133 CL15 isn't necessarily better than DDR3-1600 CL9.
Well, it is, but not nearly enough to quibble over. We were quibbling over him using the most outlandish benchmarks I've ever in my life seen, in an attempt to prove how superior the Skylake architecture happens to be. Did you not read the page or 2 before you posted?
Here's his first attempt:
http://www.eurogamer.net/articles/digitalfoundry-2015-intel-skylake-core-i5-6600k-review That's some shill using a 1.2 Ghz overclocked 6600k, along with a +25% overclocked memory controller, compared to a bone-stock i7-3770k and 4790k. The shill never even mentions what speed RAM he used with the IVB and HW, hence my "or even 1,333 Mhz DDR3" comment.
WTH? That costs $190 here, no shipping charges, no taxes/VAT:
http://www.newegg.com/Product/Product.aspx?Item=N82E16820144756