I'd be pretty happy if governments all decided "cyberwarfare" was an acceptable substitute for the real thing. Nobody dies, some money is lost, some important people are embarrassed.
In short, critical infrastructure all over the world is being needlessly put at risk. The operators are placing their own knives across their own throats, and all a network-based attacker would need to do is jog a few elbows.
A few people might die. It's not like anybody is putting kill-bots out there, connected to the Internet, with an easily-toggled BERSERK_RAMPAGE flag (yet), but I think water treatment plants and electrical power grids are probably vulnerable to attack, and could cause some folks at the margins to die.
If there's one thing the Sony breach and other major hacks over the last few years have illustrated, it's that there is no clear limit to the damage that can be done from something like this.
EDIT: To me, what's so frightening about this is the lack of historical precedent for conflicts between nations in the form of cyberwarfare. We don't know what could happen. We don't know if the damage will be limited to financial cost and embarrassment, and we don't know that a scuffle between nations on the Internet will remain contained as such.
People would definitely die.
Which is pretty much exactly why it won't happen.