Is North Korea's talk of war serious or just bluster?
One advantage of North Korea's crude bluster is that it dulls reactions in South Korea and the Pentagon. Oh, there they go, again! But if North Korea intends to attack, this tendency will buy them time to mobilize.
During the Cold War, if I recall correctly, we generally assumed the Soviets would be able to get a week head start on NATO in mobilizing for war. It would take us that long to recognize that Moscow was seriously preparing to lunge west and to make our own decisions to prepare for the onslaught.
It may depend on how good our surveillance is these days. No doubt we look at a lot of indicators to see if enough are tripped to indicate North Korea is preparing for war. While North Korea's army is mostly forward deployed near the DMZ, it can't roll out of the barracks on no notice.
But do the North Koreans know what we monitor? If they do, can they disguise mobilization signs long enough to get a week's advantage on South Korea?
I know, why even ask this when North Korea would lose a war with South Korea and America?
Because I don't know if North Korea believes they would lose a war with South Korea and America. Or if they do believe that they'd lose a full scale war, I don't know if the North Koreans believe that taking military action against South Korea, Japan, or even America means that a short and glorious attack must escalate to a war. The North Koreans might believe that their old strategy of rattling sabers and killing South Koreans in small numbers to inspire fear and money still works, but that they only have to up their game a bit and draw some serious blood--or some new blood. A few missiles aimed at Japan or Guam might be just the trick to regain that fear that North Korea counts on for money.
One thing we should be grateful for is that North Korea lacks much ability to attack American soil (other than bits of Alaska and Guam). North Korea's gleeful threats to target American cities surely telegraphs their intent and likely their belief that the ability to target cities in North America gives them leverage over us.
And you never can tell if the North Koreans believe they would lose a war. Seoul is enticingly close to the DMZ, after all. It starts the war within range of some of North Korea's longest ranged artillery. North Korea may believe that a combination of bombarding Seoul and massive infiltration by North Korea's large force of special forces (not SEAL or Delta quality, I should add--think more like approaching Army Ranger quality without the technology) will sow enough confusion in the ranks of South Korea's army that even the North Korean army could slice through them for the short march to Seoul.
Take that city and North Korea can call it a victory and call for negotiations to refrain from destroying the city and killing as many of the people who failed to run south as they can. What would South Korea pay to get their capital (where a quarter of the population calls home) back?
If you play the odds, North Korea is blustering and bluffing as they always have, even though the bluster is more specific and legalistic (in nullifying the ceasefire and proclaiming they are at war) than in the past. But in theory, such bluster can cover mobilization if you intend to attack and simply want to maximize your mobilization advantage.
How good is our intelligence in monitoring relevant North Korean military activities?