Point the first, almost all software projects are mismanaged and under-resourced.
Point the second, in my experience and the experiences of many friends and colleagues in or adjacent to software development (so: here come anecdotes, but I'd like to submit the are representative) -- the loss of two key people is ALWAYS crippling on ANY team in ANY size company on ANY project REGARDLESS of how well managed and funded it is.
We might like to think that large organizations have in place redundant backups and knowledge transfers and blah blah blah and some of that is true, but the reality is that software is still build by individuals working in small teams (or often: working solo). Thus when you lose that one guy or gal it is a BIG PROBLEM.
I will give you an example from my current employer, which is a huge multi-BILLION dollar company you almost certainly have heard of. We have a process that sends some key product information to our suppliers (trying to be vague); that process and much of the code was written by one woman. She is still, thankfully, with the company and not dead like the tragedy that befell the Wizards team -- but she is off in another area and this is, frankly, no longer her problem. It has been a COMPLETE DISASTER since she changed teams, as the new developers try to figure out what she did (I dunno, maybe it's junk? it worked, though), the new non-development manager tries to answer questions from the supplier that he can't answer, and everyone is pointing fingers at everyone else.
This kind of thing happens ROUTINELY at major, very well capitalized companies doing work far more serious than elfgaming.
Thus, how could Wizards be immune to it?
So when someone says "part of 4e's failure was their crappy software development!" that complaint applies to LITERALLY EVERY BUSINESS IN AMERICA.
Professional software developer since 2002 here.
In (slight) defense of my profession, I probably should point out that software DOES somehow successfully get released from time to time! Looking back, I can only think of a very small number of projects I've ever worked on that could have been derailed by the loss of two people, and this is in teams numbering 5 to 90, in tiny family businesses or major banks. IF you have someone who is a knowledge silo, you break your back to share that knowledge - you assign someone to shadow them, you assign other people their work in order to share the experience, you make them run training sessions on the material, you hire in experts.
A bit more in the weeds though. The 4e books were released in June 2008. I'm a little fuzzy on recollection, but I vaguely recall that the initial intent was for the online tools to be available at release time (please correct me if I'm wrong). The murder took place in July 2008, a month after the release of the 4e paper books. The application should have almost been ready for release at this time, but instead, it was in such a state is was utterly abandoned, with WotC/Hasbro believing that it was so hopelessly irretrievable that they simply wrote off the no-doubt-significant sunk costs. At this tie they should have been deep into fixing the final few bugs, doing last-stage testing of scaling, security, performance, etc, writing documentation and tutorials. If they were at that stage, then the murder would not have resulted in the cancellation of the project. Delay sure, but the thing would have been close enough to finished to be shipped eventually. But this was clearly not the case.
Now, as I know VERY well, production delays happen, deadlines slip, and I can't imagine that, given later tragic events, the development team had been working at prime efficiency for some time.
But.
A project doesn't get into this sort of state overnight. If, on the target release date, your code is still so incomplete that the companty junks it all rather than even try to finish, the problems go back a year, at least. Where was the project manager during this process? Why hadn't additional people been brought on 12, 18 months earlier? (Yes, I know more bodies aren't a cure-all, but the project was clearly just SO far behind that they were basically a necessity, although not the only one by a long shot) Why weren't they aware that 1-2 people were irreplaceable knowledge silos and take immediate steps to fix that? Why weren't the WotC decisionmakers getting frequent updates and demos of the software, so they could twig that something was wrong? Hell, once it became clear that the software wouldn't be ready in time for the paper release, why didn't they bring in an expensive top-of-the-line sort of hired gun to try to push the thing out in a hurry? I mean, this is D&D for freak's sake - you're telling me that you couldn't find a few software geeks who'd jump at the chance?
There is a level of very basic project management competence on display here. Now, I admit I've long forgotten whether this software was developed in-house by WotC employees or whether they contracted out. If the latter, it's kinda more explicable. Subcontractors fib about stuff all the time to clients, although WotC should have had someone on site full-time if they were even remotely legit (they did with BG3 I'm pretty sure).
This project was a very big deal for WotC - I mean, how often do you release a new D&D edition, and you've been promising the online tools since day 1? But it seems to have been treated with almost neglect. I sometimes wonder whether the computer side of things was pushed on an unwilling WotC development team by WotC/Hasbro higher-ups, and then deprived of resources by people who thought that the paper game should have been the focus. This possibility might explain what happened, but of course it's complete speculation on my part.
Regardless, this whole project was very clearly in deep, deep trouble long before the murder, and the loss of personnel was only a small part of why it failed.