Initial comments 
Looks like a great start to me. I agree with most everything you said here already, got lots more to add.
I slightly disagree with the Remove button placement but that's because we want to be able to remove items "in progress", and in non-HTML5 browsers that's where the spinner goes. So the close box should probably go somewhere else -- perhaps appear on hover to the top left outside the upload box. Or, the other alternative is to leave it where it is but on another line below the soon-to-be-added progress bar.
Also, the really correct interface would simply start the upload the instant it was added, which changes things somewhat. There shouldn't be a "upload" button.
One correction though; our goal is not more uploads, but more resources for the world. Hence the need for both ease of use AND user education. We should not ignore the experience of Commons volunteers, that explaining this concept to people is difficult. Most people think that anything they find on the web is "public domain".
NeilK 20:10, 30 March 2011 (UTC)
Tutorial dismissal 
Agree with Neil, lots of good ideas here. But, I disagree with the changes in the "Learn" section. The cartoon is really an essential learning device which we want people to notice (as opposed to accidentally clicking it away). Both suggested changes (easier continuation/default dismissal) are IMO likely to dramatically reduce the impact of the tutorial.
So, with regard to:
The primary goal of the user here is to upload files. Requiring them to "go to school" is counter-productive to this goal.
Actually, with Wikimedia Commons, "going to school" is really an unavoidable part of the experience (copyright is hard, and Wikimedia Commons is a free image repository), and is much preferable to getting your files deleted or having an unpleasant experience later. Certainly we will need to try varied approaches here to see which ones are actually effective. But, we definitely do not want to encourage folks to just pass through this stage of the process.--Eloquence 00:19, 31 March 2011 (UTC)
- I agree mostly with Eloquence here, but we have not actually tested whether the current tutorial results in the education we desire. It is imaginable that Brandon's approach might actually work better. Especially the "donate" wording, which Gotomedia independently suggested. I propose that we don't argue about this at all, but try to get an answer for this with some simple hallway usability testing. NeilK 00:27, 31 March 2011 (UTC)
- Yes, various forms of testing effectiveness would be great. To the extent that we'll have to make intuitive choices, we'll have to pick whichever approach we think is most likely to lead to effective user education over which is most likely to not get in the way.--Eloquence 00:36, 31 March 2011 (UTC)
Donate media 
I like the idea of referring to media uploads as "donation", but I would strongly suggest using both terms on the screen. "Want to donate files to the Commons? Select media files to upload." Instead of "Want to donate.. Select files to donate". We don't want people to have any misunderstanding about the transaction, and while "donation" can be a helpful framing device, I think the term "upload" should occur in the transactional context (the upload buttons and any subsequent UI messages referring to the transaction).--Eloquence 00:23, 31 March 2011 (UTC)
Tutorial format 
I really like this flowchart: Should I post this image on my blog?
Something like this might be kind of fun to put together -- I imagine a sort of small window showing a portion of a bigger "poster" where clicking on arrows scrolls to the next thing. That wouldn't have the below-the-fold problem. And if you look carefully, it's all text even if it looks like it's hand-drawn. So that makes it easy-ish to internationalize. And maybe even fun to read.