Nice write-up and in the end everyone needs to admit that email is still the way most people go; even if it is just to tell them that they need to log in to the online collab tool because there has something changed.
One question I have though is how far you have to go with -not- preventing user's mistakes. I miss some concrete examples (except for the Little Britain sketch).
in the end everyone needs to admit that email is still the way most people go
I think this is right. For us, it's collaboration tool and email, not collaboration tool or email.
We're using email + Google Docs -- pretty barebones, but the basic mechanisms work. This is for a loose, geographically distributed, sometimes amorphous group that is very resistant to adopting and converging on new technology tools -- not due to technophobia-- people are too busy doing experiments, writing grants, etc. Every new tool has high activation energy.
[Side comment: One reason I don't like non-email collaboration tools because they make it hard to own my history. With email, I have my own copy of what I was told and what I said. I'll have it after I leave the job; I have it when I'm offline. I can search it. I can do what I want with it. It won't disappear when the hosting company goes offline, or the hosting tool becomes obsolete. I don't think other people care about this as much as I do...]
- Construction collaboration is all about official issues of files (i.e. "I issue these 10 files to you for comment"). We make it easy to fix an issue after it's happened, if there were errors in the issue.
- If you're invited to the wrong company, you can easily be moved to the right company
- If someone named your company incorrectly when they invited you, you can rename your company
One question I have though is how far you have to go with -not- preventing user's mistakes. I miss some concrete examples (except for the Little Britain sketch).