For those who got stung by the bug, I really can't speak to what happens next, as I'm way out of my depth in that field -- less "I can't tell you" and more "I don't know." I'd work with customer service, as needed. (Rote answer, I know.) My role in all this was just working on the fix: I was part of the roving band of programmers last night, moving from desk to desk, trying all sorts of things to get a repro, a fix, and a verification last night. (Testing required multiple clients, sometimes from desks on different ends of the building, so I was literally dashing around the halls last night. Got my cardio in, at least.)
As for what actually happened, it's...complicated. A collection of disparate things, really. Involving a very sick dev (who was out of commission that day), risks stemming from setting up our tech infrastructure to support development of a steady flow of new shinies on a live game, and, frankly, a really, really big code base. I don't mean that as a weird humble brag -- those are the actual ingredients here. I've seen some conjecture that we forgot to copy-paste code from PvE into PvP / WvW, or that we don't care about those modes, or whatever. Let me be the first to assure anyone feeling that way: neither is true. (And that's just not how things work here. :P)
For the more nuts-and-boltsy among y'all, some deets: basically, the crash required a set of circumstances that just don't converge except in content with lots (not few) of human-controlled agents (not NPCs) that are customized for aesthetics (not re-created freshly, as is often the case in development for the purposes of controlling variables). Translation: the crash really only happened reliably on Live (ruh roh) in player-vs.-player contexts. Which meant that WvW, PvP, and the demo derby took it on the chin. /trombone
TLDR: We feel ya, WvW and PvP! I promise promise promise, there's nothing underhanded here. Just a bug. Eventually squashed by a stampede of programmers late last night.