Other sites are reporting this is due to Full Disk Encryption. A Nexus 6 was tested with and without FDE and with it was slowed to a crawl. It is up to the supplier as to whether is it turned on and in some cases there is no option to turn it off. No 5 for my M8 until I read reviews on this matter.
Yeah.
And the prerelease AOSP that Anandtech - the actual source of all this - was found to not be using the Qualcomm (maker of the Nexii processor and yours) hardware functions specifically for storage I/O.
Regardless, a whole thread exists at XDA with the conclusion that despite the AnTuTu benchmark results, no actual performance difference is felt by users.
Read up all you want, information is power but don't mix apples and oranges.
In 2012, the big Nexus fail on storage was dirty buffers due to an error in the kernel build procedure.
Despite all warnings to the contrary, many HTC owners without the problem ignored the patch warning to apply only where needed (HTC never got that wrong despite the beliefs of the Nexus Faithful), applied that patch and promptly bricked their HTCs.
You can't read about prerelease or even release Nexus failures and then go along with the masses that it proves something about the latest Android.
OK, you can. But if you believe it you're going to brick a non Nexus.
Nexus 5 and Nexus 6 and Nexus 10 issues have no more to do with the HTC One M8 than the Droid Turbo.
Unfortunately, every blog you're going to read will tell you otherwise.
I've done Android rom and kernel development, and kernel development for Linux and several BSD variants. I know the truth OK.
Nexus kernel - is not any other kernel.
It's not the same from Nexus to Nexus either.
The M8 is likely to have some 5.0 bugs. Coming later and coming with an HTC kernel - they're going to be different.