Until fairly recently I was working for a fairly major publisher (not one of the big five, but in the top twelve in the United States) in their ebooks division.
Between this article and the one at the New Republic, there are way too many things for me to comment on, but I will try to make a couple of points.
First off, I have absolutely no idea how the New Republic arrived at a 75% profit margin for each ebook sold. The main reason that this number sounds bunk to me is because the margin on every single title is vastly different and the median margin is way, way lower than that. If that's an average (and even then it sounds utterly outrageous), then they're not understanding the place of best sellers (mega hits) in the publishing world.
Also, "profit margin" more than suggests that these are "profits." They're not, because NR is obviously only factoring in the one time production costs of an ebook and ignoring the massive costs of actually running a publishing company with sales and marketing & support staff, infrastructure, and the massive cost of publishing books that never earn out their advance. I should also say, related to this, I was involved closely in the production of ebooks, and the actual creation of each ebook still requires thousands of dollars. They're not as cheap or fast to produce as even my own managers used to expect (with the casual, "why do we even pay you?" arrogance when I couldn't create a book in two days for some Amazon promotion that they wanted to join).
Second, about bestsellers, they do make publishers a lot of money. We had four titles at my publishing house that were bringing in a gross of $24 million a year some 30 years after they were first printed. However, those kinds of books are insanely rare.
I already mentioned this, but most books don't earn out their advances. Even if a book is phenomenal, that doesn't guarantee that it sells well. The print book world is built on forecasting: you need to predict as closely as possible how many books a specific novel will sell, and then print that many. If you guess under, you'll sell out and Amazon/bookstores will end up getting pissed because you can't fill orders. Too many and bookstores will return books to you and you have to eat production and shipping costs for those books. I was very glad that ebooks aren't like this, except that I was continually producing books that would sell twenty copies a year when I had just sunk $2500 into making them. Generally, that book is going to take years to pay off.
The big books allow publishers to make the small books. And take risks with some authors that try new things.
Third, most books (and ebooks) don't actually sell at the cover price. There are always daily deals on Amazon, B&N, or Apple and sometimes they're ridiculous. For example, one of the e-cookbooks that I worked on retails in print (and originally in e) for $40, and the ebook went on a weekend sale for $1.99! And even with that, we only sold about 5,000 copies. It bumped us up in the rankings and we eventually sold another 1,000 copies at the full print cover price (this is super secret publishing strategy, everyone).
But that means that the average sale price of a $40 book was $8.33 per book over that retail period. This specific case was ridiculous, but this happens a lot on a smaller level. On an earnings report you see the list price, and the average price per unit sold. I think the best case scenario I saw on a year was about 90%, and that was the best case. I would hazard to guess that most are lucky to do 75%. And big promotions often result in 25-40% averages.
Fourth, about the Apple price fixing thing, the New Republic article mentions this in the last paragraph, but that was actually an attempt to prevent Amazon from becoming a monopoly. Apple and the big five lost. Yeah, it was about higher prices for consumers, but I think consumers might have won a short term victory at a long term cost.
Fifth, the cost of ebooks and writer's share: Yeah, I disagree with the publishers on this. I think a $9.99 ebook is a fair price (none of this $26.99 for an ebook crap) and I think authors should be making way more in e-royalties than they currently are. Most authors I worked with had an e-royalty of no more than 15%. Given a book sells a certain number, say a thousand copies, I think that it should go up to 25% or more. Publishers aren't handling these aspects particularly well.