Unfortunately, everything else turned out to be accurate. Tor's official press release can be read here, editor Harriet MacDougall's comments are here and writer Brandon Sanderson has a lengthy and eye-opening essay on the situation here. So far, Orbit have not commented on the situation and the plans for UK publication of the book remain unclear.
The fanbase's reaction to the news has been mixed and although Brandon's very well-written piece has mollified opinions somewhat, it simultaneously revealed some very odd messages coming from Tor about their reasoning for the book to be split into three, rather the previously-proposed two.
The primary reason for the split is because Tor wanted to get a Wheel of Time book out in 2009: the previous volume Knife of Dreams, had been published in 2005. This is no doubt laudable on one level, as the previous biggest gap in the series had been three years, with two more common. Keeping the gap to a minimum is sensible. Tor's other excuse that if it had been more than four years than the series' profile and sales would have suffered can be immediately and summarily dismissed as utter nonsense. A Song of Ice and Fire's sales and profile grew substantially in the five year publishing gap between A Storm of Swords and A Feast for Crows. The gaps in Jean M. Auel's Earth's Children series have reached decade-long proportions and new books in the series continue to be immediate bestsellers. It's been six years since The Da Vinci Code came out and Dan Brown's next book will likely still be an immediate hit.
As Sanderson explains in depth, if the two-volume solution had been pursued, than A Memory of Light Vol I would not have come out until the spring or summer of 2010, followed by the second half 12-18 months later. He also explains that this is because the punishing pace he has set himself for the last year and a half has left him creatively exhausted, and he also now needs to break off for a time to work on two of his own projects that are under contract to other publishers. That is fully understandable and I doubt many people would complain at this. Also, given the wait that has already elapsed for this book, I doubt many fans would have balked at learning there was another 3-6 months to wait for the first book. This is not a huge deal to the fans. In fact, splitting the book in three has entailed more work for Brandon, since each book needs additional material to make it as self-contained as possible (according to the author, 25,000 words had to be added to Volume I solely to make it as independent an entity as possible, and more will be required for Volume II), and could extend the overall size of the project to something close to 900,000 words, which is a truly vast amount (the eleven published volumes and the prequel combine to form 3.4 million words).
That the fans would wait an extra few months for the two-volume option is obviously a given, but Tor continued to issue bizarre reasons why the book had to be split in three. They lamented that booksellers with their limited shelf space would be unhappy with a big, thick fantasy book and this wasn't the 1990s any more when such books were common. They also claimed that printing and binding would make such sized books economically unviable, which is a claim that had me blinking at its self-evident absurdity.
Tor is the publisher of all the previous Wheel of Time books in the USA. The two-volume solution for A Memory of Light had each book coming in at between 350,000 and 400,000 words. Those are big books, no doubt about it, but not unreasonably so. In fact, the previous longest books in the Wheel of Time series itself were The Shadow Rising (at 393,823 words) and Lord of Chaos (at 389,264), and each just about scraped over the 1,000 page mark in mass-market paperback (the normal cut-off point for most publishers is somewhere between 1,200 and 1,300 pages in mmpb for large-selling authors). So this size of book is not unprecedented in this series, let alone the genre or for the publisher.
Tor claim that it isn't ten years ago anymore and publishing such-sized books is difficult. Well, those earlier Wheel of Time books remain in print. Let's assume that older books are cheaper to keep in print than newer ones (replenishing backstock requires much smaller and therefore cheaper print runs than publishing a new book from scratch) and that just because Tor can keep pumping out new copies of The Shadow Rising regularly, that doesn't mean they can publish a new book of the same size (which would require a much vaster print run). Except of course they provably can. Steven Erikson's eighth Malazan novel, Toll the Hounds, will be published in mass-market paperback on 4 August this year. The book will be 1,280 pages long in mmpb (for those counting, that is almost 300 pages longer than The Shadow Rising). I'm not sure of the word count, but it clearly tops 400,000 words with change left over (unless Tor are using some whacked-out typeface and margin sizes). Erikson is a traditionally low-selling author who's now making some headway, but still the most generous description I've seen of his status is 'midlist'. His American sales are a barely appreciable fraction of Jordan's, yet Tor curiously seem able to keep his books in print in both hardcover and brick-sized mass-market paperbacks, and are still doing so in this current 'difficult economic climate' (a great buzz-phrase being used by many companies to justify steps that involve milking the customer for every penny they can).
So, the problem clearly is not the two-volume option's anticipated size, nor is it printing and binding practices. Tor themselves are printing books considerably larger than the books in the two-volume option as one novel with no problem, at this current time and for release this year.
This of course also by extension dismisses the bookshelf argument. Bookstores are complaining about big books taking up shelf space they could be spent on other, smaller, better-selling and more profitable books. This is especially notable in SF, where I am informed the US bookstores have seen significant section shrinkage in recent years. Again, all of that would be fine if it wasn't for the fact that Tor themselves are publishing 1,300-page paperbacks and presumably selling them to stores (if the stores weren't ordering them, the books wouldn't be selling and Tor would presumably be dropping the series or splitting the books, which they are not) at this time. In addition, the Wheel of Time series is one of the most recognisable and marketable franchises in the entire SF&F genre, a guaranteed seller with no less than four #1 New York Times bestsellers to its name. No, bookstores would clearly stock and sell this book, especially as the penultimate book in the series would be extremely marketable as well and could lead to improved sales of the prior eleven novels before it.
At this point we have to switch to supposition, which readily provides two answers. The first is purely financial. The best time to bring a book out is October-November, just before the Christmas rush (but not during it, at which time it would get swamped and lost). Bringing out one book for Christmas 2009, 2010 and 2011 is clearly a better and far more profitable option than bringing out one in summer 2010 and another a year later. Tor are not doing badly at the moment, but we have seen apparently watertight companies toppling like dominoes in this recession and it would be essential to our field to keep the USA's biggest SF&F publisher going. Tor shoring up their finances in this manner and giving them a healthy balance sheet for the 2009 financial year could be extremely important to them and to the wider genre. From the POV of someone working for Tor, this course of action is even understandable and a safe bet to take.
More convincing still is the creative argument. Brandon Sanderon's first estimate was that this book would be 200,000 words. He almost immediately raised that to 400,000, and very quickly put it up to 600,000 and then 650 and then 700 and now 800,000 words. There is no reason to think that, even with the best will in the world, the book might not still expand further. With two 400,000-word volumes, there isn't much room to maneuver if the author decides he needs another 50,000 words to do justice to the story. By expanding the novel to three volumes, the author suddenly gains a fairly substantial amount of leg room if he needs the second and third volumes to be bigger than previously thought, and if he doesn't then 250,000 words is still a very substantial novel (Joe Abercrombie's Best Served Cold is about that long, for example).
The problem, and the reason why a lot of fans are angry with this decision, is that these answers and explanations were not offered by Tor. Instead, they attempted to hide behind marketing-speak and unconvincing spin, and to do that to the fanbase of their flagship series (Tor is sometimes called the house that The Wheel of Time built, and the success of that series has subsidised them taking a chance on many other first-time authors over the years) is pretty low.
Whatever the reasoning, the situation is at least set and we know what is going on. The Wheel of Time Book 12, A Memory of Light Volume I: The Gathering Storm (try saying it whilst drunk) will be published by Tor in the United States in November 2009. I suspect that Orbit will follow suit in the UK, although they haven't issued any statements yet.