Enablin the write-caching option would indeed show higher standard HDTune benchmark scores; but people should know this benchmark produces WRONG results for RAID arrays. Its a single-queue 32KiB raw benchmark, while filesystems will have a much higher queue depth for sequential access.
If you're on a RAID, be sure to use proper benchmarking utilities. You may think that your performance sucks while in fact its just the benchmark application which is doing things so low-level synthetically that its results are useless. Filesystem performance is actual performance; including all fragmentation and filesystem optimizations.
Enabling the 'Write Caching' option in Intel ICHxR RAID drivers will enhance performance even more, but the first benchmark the OP posted is clearly wrong and this is known to the author of HDTune. But he doesn't want to change the benchmark. Sadly, lots of people think their RAID underperforms because of this and go for a slower solution instead that produces higher HDTune scores.