Hi Ville, Thanks for your reply. I don't have much more to say on this yet, but have added a few comments below.
I was in a similar situation a few years ago (writing my PhD thesis). I think you are absolutely right when you avoid Word and everything Wordish. Making a big document with Word requires a lot of knowledge about what you should avoid. And in the end you'll still spend your nights wondering why the **** the crossreferences or page numbers go wrong.
Absolutely. Word seems easy at first, but I've watched people go gray trying to get large texts to do what they want, close to deadline.
However, your wishlist looks a bit difficult.
Actually your comment here might suggest how far we have to go then, as I'd consider my wishlist a very roughly stated but really quite minimal set of requirements for academic writing.
The situation becomes much more complicated if you need RTF. It is a completely different story, a word processor editable format. I guess you don't really want to distribute your work in editable format, and PDF can be read with virtually any computer.
I'd say it'll fill number 2, as well. But RTF, no. There may be kludges to make it kind of, you know, a bit like, errr, RTFish, but nothing really good. The reason is simple: the two things are far apart from each other.
Since posting I've thought a bit more about why I wanted RTF, and realised it wouldn't do what I wanted anyway. The 'inter-operation with Word users' I was referring to is primarily this: it's common amongst academics I know here in Australia to use some of the collaboration features of Word (marginal comments and revision control, particularly). RTF wouldn't actually help with those anyway. So there's really no way around this without using Word, which I will only do at gunpoint.