They've stated in interviews that they and WotC had a different "vision" of what the product was to be. At least to start, WotC wanted to emphasize character building, whereas Trapdoor was concentrating on the content management aspects of the software, and because they couldn't find common ground, WotC pulled out.
I've stated (more than) a few times around here that it has to be more than this, because I can't imagine that a contract was signed between the two parties where it wasn't explicitly stated exactly what Trapdoor was to deliver.
One of the few things we do know is that Wizards didn't pay Trapdoor anything up front. No doubt there was some kind of revenue sharing plan for Dungeonscape, but it wasn't the usual kind of arrangement where you hire a contractor to write software for you.
Here's a hypothetical story.
I don't claim it's true. I'm sure I have at least some of the details wrong. But it's consistent with what we've heard and the way both parties have behaved.
Trapdoor had, and has, grand visions for their software. They considered it revolutionary. They pitched it to the D&D team at WotC. The D&D team was operating on a shoestring budget (over the course of 4E, their staff was sliced down to almost nothing), and were eagerly seeking licensees. Revolutionizing gaming wasn't on the agenda, they just wanted electronic tools for 5E without having to spend a lot of money on them. Trapdoor was willing to pay its own way on development costs, and that was good enough for Wizards.
So, just like they partnered with Kobold Press to make Tyranny of Dragons, they figured on partnering with Trapdoor to make what ended up being called Dungeonscape. As such, it wasn't necessary to nail down every detail of what Trapdoor was going to build. It had to have certain basic functionality, like a character generator, but beyond that they had a relatively free hand. Trapdoor was a licensee, not a contractor; their incentive to do a good job was that if they didn't, they wouldn't make any money.
However, Trapdoor was taking a long time about delivering. Months had passed since the PHB hit the shelves, and there were no e-tools, just a crappy web beta. Gamers were annoyed. Wizards was annoyed too, and met with Trapdoor to push them to get their product out the door. Trapdoor said they weren't willing to compromise their vision and to be patient, it'd all be worth it in the end. Wizards replied that they didn't care about Trapdoor's vision, they just wanted a character builder and some DM tools, and could you nail them down and ship them, please? Trapdoor wouldn't budge. Wizards decided they no longer trusted Trapdoor to deliver, so they pulled out and set about developing their e-tools some other way. (Another possibility, not mutually exclusive with the first, is Trapdoor asked for money to finish development and Wizards said no.) Trapdoor then turned to Pathfinder as the obvious alternative to salvage the project.
Like I said, this is just one possible story based on the public information released to date, and I have no doubt at least parts of it are wrong. But I'd be willing to bet that's the rough shape of what happened. Whether Wizards was right to pull out or not, is a question of whether the software is as revolutionary as Trapdoor says and whether they can deliver on it. My instinct is to be skeptical, and I've criticized them pretty harshly over their claims. But apparently they did have
something really amazing to show at Gen Con. The real shame would be if they do in fact have something revolutionary, but it fails because they can't figure out how to get it to market given their constraints. Wouldn't be the first time that's happened, or the hundredth.