This is going to be a relatively short "Crib Sheet" piece: turns out I've already written most of it:
Here are my memories of the history of "Iron Sunrise", including an excised section with a talking cat sidekick.
Here's an explanation of the structural problems with the Eschaton universe which blocked me from writing of a third book in the series. (Note that some of the ideas for "Space Pirates of KPMG" eventually surfaced in a highly modified form, in different space opera project of mine — Neptune's Brood, which is due out early this July. Which, ahem has a fan.)
Now for a meta-note: I made a contractual mistake common to first-time novelists with "Singularity Sky" and "Iron Sunrise", which I shall talk about below the fold, in case any other first-time novelists are reading these notes ...
When you sell your first novel to a major publisher, you will discover that they are not buying your first novel. Nobody knows who the hell you are (unless you're William Gibson) so it will probably disappear, and certainly won't be a runaway success. But the canny editor knows that if your subsequent novels are up to the same standard as the first, you will gradually acquire a following and begin to show a net profit around the second or third book. So rather than issuing a contract for just one book, they'll offer you a two (or even three) book contract, with an option (right of first refusal) on your next novel thereafter. Because what they're actually buying is a stake in your subsequent career.
These days, editors are workflow/production managers who work for (or at best, alongside) an all-powerful marketing department. They do not have the authority to sign off on an advance for a book (which may in a few cases be larger than their own annual salary, and is almost always more than a month's wages) without some oversight; in particular, there's no point buying a book if marketing think it's a turkey (because if the marketing folks lack confidence in the product, it almost certainly won't sell). So proposed acquisitions have to run the gauntlet of an internal committee meeting in which editorial, marketing, production, and finance all have to agree it makes sense to issue the contract.
It is much easier to convince such a committee that book #2 from an author is a safe bet if it is a sequel to book #1, because it is much easier to describe a sequel to an already-written book than an unwritten random-other-novel. So the contract offered to a first timer will typically be for Book One [insert title of the book that $AUTHOR is selling] "and a sequel, Book Two [title to be decided]".
Now, it's not hard to break out of this trap if you already have a chunk of book #2 written. If I'd said, "Book Two will be 'Iron Sunrise', a different and better space opera, and here's the first four chapters and an outline", my agent could almost certainly have decoupled "Iron Sunrise" from "Singularity Sky" in the contract — and, more importantly, given me a clean sheet to work with, rather than having to work with the design errors built into the universe of "Singularity Sky".
Unfortunately, at that time I still thought of "Iron Sunrise" as being book 2 in a series—rather than "Singularity Sky" as being a prototype for a space opera series, and "Iron Sunrise" as being the first book in the series I should have written. So I didn't object to the contract for "Singularity Sky and a sequel", and I wrote a sequel ... then subsequently tripped over my own first-timer mistakes.
The moral of this story is that if you're a first-time novelist, it's a mistake to assume that your first novel is the first book in a series. It's a first novel: there are probably flaws that you won't notice until you've written a couple more. Publisher try to lock you into a series because it's convenient for marketing purposes, not because it's mandatory: try to leave yourself some maneuvering room.
Finally, in 2007 I began writing "Saturn's Children". I was under contract to write a space opera: originally it was meant (per contract) to be that impossible third Eschaton novel. When I realized I couldn't do that, I didn't set out to design a multi-book space opera universe—but I tried to make the universe of "Saturn's Children" internally consistent. That helped a lot. A year after it came out, Jonathan Strahan approached me for a short story for a hard-SF anthology: I came up with a short piece set in the same universe as "Saturn's Children" and realized that, yes, this universe was sufficiently soundly constructed that I could set more stories in it without breaking things at random. Which is why "Neptune's Brood", due this July, was set in the same universe. (I have some reservations about the ending of "Neptune's Brood" which, arguably, changes everything—but it's a single-point change, and I don't think it introduces paradoxes in the setting that preclude further stories. In fact, it may even be a launchpad for other stories in its own right.)