People here are mixing two separate issues into one.
Issue 1: Advanced format (4k sectors)
All new models of hard drives that will come out will be Advanced format, it's a necessity in order to support the high capacities that are coming now and in the future.
Currently all AF drives support emulating 512Byte sectors for operating systems that don't support 4k sectors. WD using jumpers, Samsung does it automatically.
The only problem this is causing is performance drop when writing to the disk. This drop is not a problem for the current generation of D-Link NAS'es. They are limited by their CPU to much lower write speeds.
This issue has no influence on whether a disk is suitable for RAID or not.
Issue 2: Desktop drives not supported in RAID
All disk manufacturers have a separate line of disks that are supported for use in RAID configuration. These drives have the fabled TLER, ERC, CCTL enabled, and may be manufactured to better standards (nobody knows). Where I live they cost more than double what the desktop drives cost, so it's not strange that I will try to use a desktop drive in my home setup.
It's also not strange that the disk manufacturers will say that their desktop drives are not supported in RAID arrays. They want the higher profit margin from their professional customers. If desktop drives had TLER and was working fine in RAID arrays for home users then their pro customers might try them as well (since nobody knows if there really is any hardware difference between pro/raid disks and desktop disks)
So what's the problem?
When a desktop HDD encounter an error when reading/writing it will keep trying for a considerable time, and will then try to reallocate the sector to another place on the disk. If it succeeds there will be no error for me as a PC user, I will not be aware that anything happened.
High performance professional hardware RAID controllers don't want any of that.
If a read/write error occurs the RAID controller wants to handle that. It can recreate the bad sector from the other disks in the array, and will reallocate it without help from the disk.
The delay that the desktop drive does when correcting the error can instead cause the RAID controller to remove the disk from the array, as it thinks the disk has died.
Home NAS'es like the D-Link ones don't have to behave like the pro hardware controllers. The D-Links do RAID in software, and we users have other priorities than lightning quick error recovery. I would rather wait a minute while the NAS is fixing a read error than have the RAID array degraded.
I hope the software developers behind my D-Link NAS have adapted the RAID software to account for the behaviour of desktop drives. It would be stupid not to, but I don't know if they do...