One of the problems with properly documenting the first year or so is that the people involved mostly don't want to talk about it, or want to turn it into the stuff of legend.
The one person who was there for the full first 5 years besides Jeff (and Mackenzie, to whatever extent she was deeply involved, which I never sensed was a lot, but I may be wrong about that) doesn't like talking about Amazon publicly (though he did a little for a recent PBS documentary). Lots of other people came and went during that time (including me), and we're all a bit like the blind folk feeling out the elephant: we know part of the story, but not the full picture.
Jeff himself doesn't seem very interested in talking about the early period, or if he does, it seems important to him that the story describes a trajectory that obviously connects with what was to come (which is not entirely unfair).
The people who stayed for the long haul generally don't want to talk, it seems; the people who left mostly have not-so-positive stories to tell at this point (partly because of the behemoth that the company has become).
Personally, I think that Amazon's complete failure to take the high road in creating the model of retail in the 21st century is the more important story than details from the early days. The appalling treatment of employees, the conflicts between operating a marketplace and being a seller within that marketplace, and the general emphasis on our lives as consumers instead of the more important roles of citizen and employee .... this is what's really important about the Amazon story in my opinion.
Oh wait, it's HN. You want to hear about how I pushed and pulled C++ objects to and from Oracle in 1994, and how integrated that into NSAPI, amirite? :)