But that wasn't what actually caused the USSR to collapse, and there was still a window where we could have done a major, mainland invasion and prevented them from accumulating the absolutely absurd stockpile of nuclear weapons which Russia possesses today. We potentially have a situation like that today.
I don't think North Korea will accumulate a large nuclear stockpile, on the grounds that they do not have the human or natural resources to do so. Their program has progressed at a snail's pace, and recent tests still fail miserably. It's entirely conceivable that the military will overthrow the Juche regime before they ever create nuclear warheads capable of being deployed on ICBMs. Then, at the reins of a populous country with a new government, they might actually be willing to incorporate into the rest of the world economy in return for reasonable nuclear restrictions. Basically, a situation like what we have with Iran.
But even if they were to get a stockpile like Russia does, would it pay off to invade them early? Imagine the tens of millions of people who would have died as a result of an all-out war with Russia. And even still, Russia hasn't deployed any of their nukes. Would it have been worth it, in hindsight?
Most North Korean infrastructure could be bombed into the ground by the US from carriers in the South China Sea in a matter of days. The question isn't is war with North Korea a good idea, as I believe it would be worth it due to the opportunity to put a halt to the NK nuclear program.
The question is more one of "how many US/civilian casualties are acceptable to accomplish this?" Additionally, "How do we protect South Korea/Japan from retaliation, particularly nuclear retaliation by NK once they realize it's over?"
Knowing Mattis, he's been planning this out for a while. It will be difficult to take out North Korean infrastructure and internal supply chains without killing
some civilians. Surgical strikes and taking advantage of American air superiority is obviously the ideal option, but bombs are still imprecise and civilians will inevitably die.
In other words,
civilians will die. It's nearly inevitable due to the imprecision of airstrikes that some will be caught in the crossfire. The question isn't one of "Is war with NK to stop their nuclear program from progressing worth it?" because IMO it totally is. The objective should be to neuter the regime as quickly as possible while avoiding the deaths of civilians and protecting SK/Japan from North Korean retaliation.
I'm envisioning a Six Days War scenario where NK is pushed back in a matter of days with a series of surgical strikes intended to disrupt the military's supply chain and eliminate potential missile silos, artillery positions, weapons depots, and airfields. The question is how to minimize collateral damage while doing so.
re-examining your question, you're essentially asking, "Is a preemptive strike justified if it could prevent further casualties which are perceived to be inevitable later down the line?" My answer is, yes, entirely justifiable. NK would nuke NYC in a heartbeat if they could.