[mgmueller]

4 hours (standby, no usage after charging): Still showing 100%.
12 hours: 97% (16,67 days linearly extrapolated)
14 hours: 96% (14,58 days linearly extrapolated)
18.5 hours: 94% (12,85 days linearly extrapolated)

From 16,67 days constantly going down, now to 12,85 days.
Obviously, it's not linear (be it the meter or the battery itself). So linear extrapolation obviously has an error of at least +/- 25%.
So I'll go on, way closer to 0%.

24 hours: 92% (12,5 days linearly extrapolated)
31 hours: 90% (12,92 days linearly extrapolated)
39 hours: 87% (12,5 days linearly extrapolated)

Considering a "drop" I expect in the last 3rd, it looks like less than 2 weeks. But way more than my 4 days of course.
In my poll, most users state 1 week. Does seem about right for the moment.
Let's see, whether I'll have a surprising drop in the next few days...

43 hours: 86% (12,80 days linearly extrapolated)
49 hours: 84% (12,76 days linearly extrapolated)
73 hours: 77% (13,22 days linearly extrapolated)
86 hours: 73% (13,27 days linearly extrapolated)

Hmm, that's all perfectly in line with the rule of thumb from iRex: Close to 2 weeks.
And what most of you give in my poll: 1 to 2 weeks.
Unfortunately, I won't be apple to reproduce, what I've done exactly, before the 4 days have happened. I know which book I've read, of course. But I don't know which, when or how many other documents I may have opened, before going into standby...

95 hours: 70% (13,19 days linearly extrapolated)

EDIT: iRex found a bug, which will be fixed in the upcoming firmware update.
This bug is known to be triggered by PDBs. I only read B&N PDBs. So this obviously did happen with my reader. So I'll end my test and look forward to the firmware update.
iRex definitely gained lots of ground with me here again. Haven't seen such discussions with other manufacturers...