Standby and hdparm compatibility

[dolan.m]

I've always been trying to minimize systematic HDD seek in my TS-251+. I don't intercept the processes I know are relevant or the ones I'm not sure about. I don't waste time with processes value of which is disputable, but are executed rarely (couple of times a day). But I hate processes which are absolutely useless and keep accessing the drives regurarly, e.g. debugging records in the logs. I have set my standby timer to 30 minutes, but I don't even insist on the disks spinning down, I just like the drives' heads have some rest, when no actual work is required from them.

Now I've upgraded my disks to 2 WD Ultrastars 310 6TB. I've also upgraded from single volume to RAID1 and the TS firmware to the newest version 4.3.1400. So many things happened and I wondered what's new in the HDD insomnia. There are 2 surprising findings:
1. Most of the time, when I randomly listen to the NAS (while no actual "business" process running), I can hear the disk seeking, every second or 2. But sometimes the NAS is actualy sleeping (Power LED off, disks silent), not for very long but it happens. It seems unprobable the regular access to the disks stop for more than 30 minutes (my standby timer).
2. I tried investigating this with hdparm -C. In the past I had even written a script which checks every 5 minutes and writes the disks status to a file (on ramdrive, of course). Now I'm affraid the hdparm program is not compatible with my disks anymore. When I try hdparm -C couple of times in a row, I'm getting random-looking results
active/idle
standby
Which to me looks impossible. There is no reason the disks enter standby in such a situation (no standby command issued manually or by the standby timer).
I tried to update hdparm to newer version (via Entware). It makes no actual change with one exception: standby state is substituted with uknown. This could hint that there really is a change in drive reporting it's state. But what it really means, does anyone know?