So, I setup the Gateway laptop, plugged into the same gigabit switch as the GT70. I started doing a ping to your address you listed before, and then started downloading the new NVidia Gefore driver from NVidia (at 4.0 MB/sec, maxing out my connection). Here's the result....
Have you changed the ping priority in the killer manager ? By default the priority is Normal but you can set it on Max.
I made a test this afternoon but I didn't had the time to post the result and I don't have the neetbook now (I think my wife have closed my session anyway).
I used http://www.speedtest.net/
to overload the bandwith with a very good server (from my provider) and I ping the same server with a -n 50 param. The 50 counts cover the length of the bandwith speed test. I launch the ping test just when the bandwidth test begins.
Without the overload, I got a 10ms average trip time, 0 packet lost.
With the overload and normal priority for the ping, i got a 200 ms average trip time (with a max above 1000 ms) and 5% to 10% packet lost (after 3 tests).
With the overload and max priority for the ping, i got a 170 ms average trip time (with a max around 400 ms) and 0% packet lost (3 tests).
With the overload and normal priority for the ping and a max priority for the flash player, i got a 750 ms average trip time and 80% packet lost. And the speed test gains a little better result but not very significant as I was already near the maximum allowed by my cable provider, but the ping for the speed test passed from 15 ms to 10 ms.
For each speed test, the results was all almost the same, but I think it's will not be the case if I change the ping packet size as I used the default.
I have a netgear cable router that have QoS implemented and the test was made with the wifi with a very stable (flat) signal of -70db to -72db at 2.4 Ghz N 300 mbits.
So for me, yes, QoS and the Killer network manager can do the trick if everything is well configured and the local infrastructure can support it. The 0% packet lost is a very good example that the QoS can make the difference.
For the test in the video, it's not really a good test as even if the hardware is almost the same, the software is not. The GT70 windows installation is relatively clean. But it's not the case with the G75 from asus. If the windows installation has not been cleaned, the G75 is a calf out of the box (my reseller said and he sells a bunch of it every month).