OS6 now works on x86 Legacy WARNING: NO NTGR SUPPORT!

I have installed 6.0.8 on Ultra4 and everything seems to be fine, front page reporting is broken, but other than that everything is fine. Even iSCSI is working great ( from windows 7 clients using mutual CHAP).

I have a bit "advanced" problems - please note that they should not stop anyone from using OS6 on older gear, they are just plea for advanced help:

1) I don't seem to be able to stop the NAS from "calling" home, i ran apt-get purge on some packages as recommended earlier in this thread, but i still get this:
root@poligonas:~# netstat -tan
tcp 0 0 192.168.1.72:35191 206.16.42.179:443 ESTABLISHED

Anyone got clue how to get rid of this stuff?

2) Power efficiency of OS6, it just seems to be chewing a few watts more than 4.2.24. While i don't really care about disk turning off, i decided to dig deeper and didn't like what i've found:
After running
pidstat -t -w
i saw this:
12:52:11 AM 1673 - 1.00 0.00 readynasd
12:52:11 AM - 1673 1.00 0.00 |__readynasd
12:52:11 AM - 1734 100.00 1.00 |__readynasd
12:52:11 AM - 1735 100.00 0.00 |__readynasd
12:52:11 AM - 1736 100.00 0.00 |__readynasd
12:52:11 AM - 1737 100.00 0.00 |__readynasd
12:52:11 AM - 1738 100.00 2.00 |__readynasd
12:52:11 AM - 1740 100.00 0.00 |__readynasd
12:52:11 AM - 1741 100.00 0.00 |__readynasd
12:52:11 AM - 1749 100.00 0.00 |__readynasd
12:52:11 AM - 1750 1.00 0.00 |__readynasd
12:52:11 AM - 1761 1.00 0.00 |__readynasd
12:52:11 AM - 1938 1.00 0.00 |__readynasd

Yup, you see it right, it is readynasd waking your CPU 800 times per second and not letting it go to deeper sleep states.
And unfortunately those threads are not doing anything important either, they are using technique that has fallen out of favor in any power use aware programming for years now:
strace -r -p 1734
Process 1734 attached - interrupt to quit
0.000000 restart_syscall(<... resuming interrupted call ...>) = 0
0.006898 nanosleep({0, 10000000}, NULL) = 0
0.010366 nanosleep({0, 10000000}, NULL) = 0

Yup, that is nanosleep for 10ms burning extra electricity.

4.2.24 had none of this, looking at vmstat you could see dozens of interrupts/wakeups instead of thousand+ on OS6 box...

Question is - can we get rid of all this madness? I can't dig deeper without gdb and powertop, but maybe some Jedi will have a clue wth is going on with those 8 threads?

3) Anyone got idea why Netgear is using acpi_idle instead of intel_idle? while both are using mwait in idle, intel_idle can go deeper. I wonder if anyone tested custom kernel with intel_idle, is it working?