Governments can - at any time - stop inflating their debt-based currencies and stop spending more than they take in. They have the upper hand in that any policy changes they make will ripple through the market on a large scale much faster than any alternative currency can impact the market or have such a large scale effect.

Part of the promise of Bitcoin is that it will eventually force governments to reform their own systems simply by presenting a better alternative to people. In order to compete and retain some semblance of power, governments will have to address the market conditions and become serious about reform, if only for the sake of self-preservation. If things were really looking like all fiat was going to go over the cliff of hyperinflation due to wealth pouring into alternatives, then they would have no choice but to engage in serious reform. Any such reform will help to stabilize national currencies and make them attractive enough to stem the tide of the nightmare scenario described by the OP. People will have time to adjust.

The threat of the use of force on a grand scale by governments is largely a paper tiger. It is only useful for them to use force at the margins, otherwise it becomes counter-productive to the survival of the regime itself. Any regime that goes all out with a show of force on a large part of the population at one time is setting itself up for failure because once you go that route you lose all propaganda value and even the masses are not fooled by your claims of legitimacy. You become ripe for overthrow.

The future currency game, I think, will be more like a chess game than an actual war. There will be ground gained and lost on both sides. If you consider that politics is a marketplace of ideas, and you understand that markets tend toward equilibrium, you can predict that forces on both sides will tend to balance over time. By balance I mean that both sides are always pushing back, and that tends to create short-term stability and predictability. It doesn't mean a 50/50 balance. It just means that the tension between them produces a balance one can quantify presently and predict in the short term. Once that state has lasted for a while, a new picture emerges of who's ahead and who's behind, but both are still pushing against each other. Then the opposing forces adjust their policies accordingly. It's a mistake to extrapolate present trends out indefinitely. Things tend to only go haywire in isolated spots, not in the big picture. So you might get a crazy dictator or an insane political system for brief periods of time in limited areas, but the political market is so wide that these aberrations typically do not propagate throughout the whole world.

The nightmare scenario is of course a tyrannical world-government which is not afraid of displaying an extreme show of force as a matter of policy because it loses its inhibition and thinks it has captured the queen in the grand chessboard of world politics, but it's difficult to see anything like that lasting long term. Ultimately, all these systems rest on the support of the people. They either actively support or acquiesce. But if you push too far, even those who would otherwise acquiesce are forced by the harsh circumstances to take a stand. The only reason for people to compromise with tyranny is if they see self-preservation in it. Once it becomes clear that the survival and comfort of a certain portion of the population is not predictably certain, then you can get an explosive push-back, and that's when revolutions take place.