Thanks guys. I'll continue to dig. I have set my N800's wlan_sleep_timeout to 1000mS. I'm going to my friend's now and I'll make some Skype calls thru his network and see if that helps. I'll also look and see exactly what AP/Router he has.

When I get back here ( home ) I'll restore the wlan_sleep_timeout setting to 200mS and retest in my network (which works really well with Skype).

I have some test equipment here, so I suppose I could try to do some empirical bench testing on the differences between 200mS and 1000mS as it applies to battery drain. This could quantify the difference and make it easier to decide if the 1000mS sleep_timeout is a big negative, or minor one. There are many variables, involving the manner in which the N800 is in use so this isn't going to be easy to do this test correctly. What I'm saying is, setting the wlan_sleep_timer to 200 and measuring drain, and then recharging battery and repeating with wlan_sleep_timer at 1000mS with the N800 sitting turned-on (with WiFI connected) but idle wouldn't necessarily teach me anything useful. I'd have to generate typical network data traffic between the AP and the N800 in a consistent manner between the two tests but not the type of data traffic that wouldn't show a difference between the two settings. This isn't an easy test to set-up because what is "typical" data traffic?