Actually, there is no evidence for the claim that Windows NT malfunctioned. If there was, the Wired author would write about Windows NT malfunction. He didn't. Given that the whole Wired article was based around the "Windows NT sunk the ship" story (after all, how many Wired readers would care about a bug in some proprietary navy software they never heard about?), the only reason for not saying straight that the Windows failed was the fact that it didn't, in fact, fail.
The exact cause of the problem is already known - missing NULL check in the application that... well, that did something with the ship. This missing check caused the application to crash. The only connection between this application and Windows was that it was running under Windows. It wasn't part of Windows, and it wasn't written by Microsoft.
Even the management guy (who wasn't directly involved and didn't really have any clue about what worked and what didn't) didn't claim that Windows NT failed - he claimed that he _believes_ that choice of Windows was not a good idea. In every corporation you will find dozens of people that believe Linux is bad and you shouldn't use it for critical parts of infrastructure. Exactly the same level of "technical insight".