1) Being the party/person who started another world war; 2) Being nuked without ability to retaliate (in case the executive branch is wiped out)
Most incredibly stupid decisions were made rationally, under extreme perception of the reality.
http://en.wikipedia.org/wiki/Dead_Hand_(nuclear_war)
Its purpose was to actually two fold -- ensure the ability to launch a retaliatory attack after a "surgical" nuclear strike and (and this follows from the first) ensure that hot headed generals or future leadership wouldn't be too quick to press the trigger on the very first warning (which might be a false alarm, and these have happened in the past).
That is a pretty nice system come to think about it.
So on first ever threat the #1 thing to do is enable this system. Then decide what to do, knowing, it would be possible to retaliate.
Note, what is interesting, they didn't broadcast to the world that they had it. Which is interesting. From a game-theoretic point of view, shouldn't they have broadcasted its existence as that would have discouraged a first strike?
Though, you are right about the danger of the perception of reality. (that why it's good, some time, to go back to the first principles)
If they thought they could deliver a knock-out blow to the majority of your arsenal, they'd do it. The real question, I feel, is more to do with what the window for that blow would be:
Can they achieve sufficient synchronism across their various platforms that the bombs will all arrive at more or less the same time - preventing the destruction of some components of the system from warning others?
Can they avoid your long range systems picking up their launches several minutes out?
Is the window, from detection to response, in other words, small enough that they can hit you fast enough that you don't have time to respond?
In that arena I don't think a few seconds one way or the other, to input a more complex launch code, would make much difference. It seems to me like if they can get close enough not to be picked up at long range, if they can launch with synchronous on target timing, then they should overshoot your response time by a long way. And if they can't, then they shouldn't come anywhere near it.
2) A first strike may be intended to decapitate the leadership, decimate the enemy nuclear capabilities, destroy the entire army, or obliterate the whole country; the same goes for a second or third strike. It should also be noted that the Soviets implemented the "dead-hand" system so that any surviving installations would automatically retaliate even in the event of a successful first strike.
3) There are a variety of first strike scenarios, mostly involving SSBNs near the enemy coast, and stealth bombers since late in the cold war, as they give the shortest response window. Synchronizing a first strike would likely only be between one type of platform (i.e. all stealth bombers or SSBNs launching simultaneously); which is not too challenging.
4) First strike requires less than 15 minutes of notice, though even this may be too much. Soviets feared that the stealth bombers or a missile shield would encourage an American first strike exactly because the first would provide no response window, and the second would make any response futile.
5) Above.
6) Maybe.
This all misses the basic point, which is that MAD made nuclear strikes even more pointless than they had previously been. Simply put, what is the point of destroying the enemy, and irradiating their territory (especially their most valuable assets) with a strike? Why would you put yourself at risk to achieve a goal of no benefit to you, except relieving paranoia, under MAD?
2. A locking system in place does not make much of a difference if there is no war, and much of the deterrent factor would probably still be there. If there is a nuclear attack, well, even if you are able to respond things will look quite grim, so it doesn't really make such a great difference (and it has to be seen how easy it is for your opponent to block retaliation).