My most important point is that reasoning of the form "If only the Kaiser had been less obsessed with a strong navy, Britain might have been induced not to ally with France" is likely false. Since the 1700s, Britain's policy had always been to prevent a European hegemon - the UK's opponent changed from France to Germany when the potential hegemon changed from France (Louis XIV, Napoleon) to Germany.
That said, with the benefit of hindsight, it is obvious that Germany could not be closely allied with Austria-Hungary and Russia. Both wanted to dominate the Balkans to the exclusion of any other great power: Russia for warm water ports, AH to have a freer hand against internal dissent.
Also with the benefit of hindsight, Germany looks awfully foolish for picking AH over Russia. But even if the Reinsurance Treaty was renewed in 1890, it is unclear whether Russia would have continued to be willing to renew it over the next two decades.
But my thesis is that nations act in their own interest, regardless of internal dynamics. That is not the same as saying that nations always correctly figure out what their interests are. Britain's failure to take steps to prevent the unification of Germany in the 1850-60s is as inexplicable as Germany's choice of AH over Russia a few decades later.
I think internal dynamics play a greater role than you assume. Personalities do matter in politics. To take a current example, while little has changed about the facts between Russia and Germany of today, the relationship between those two nations has changed a lot after Merkel succeeded Schröder as chancelor, simply because Putin and Merkel don't work as well together on a personal level as Schröder and Putin did.
The standard view of Mutually Assured Distruction (MAD) is something like:
Occasionally people will reply with an argument like:
This is an anthropic argument, an attempt to handle the bias that comes from a link between outcomes and the number of people who can observe them. Imagine we were trying to figure out whether flipping "heads" was more likely than flipping "tails", but there was a coin demon that killed everyone if "tails" came up. Either we would see "heads" flipped, or we would see nothing at all. We're not able to sample from the "tails: everyone-dies" worlds. Even if the demon responds to tails by killing everyone only 40% of the time, we're still going to over-sample the happy-heads outcome.
Applying the anthropic principle here, however, requires that a failure of MAD really would have killed everyone. While it would have killed billions, and made major parts of the world uninhabitable, still many people would have survived. [1] How much would we have rebuilt? What would be the population now? If the cold war had gone hot and the US and USSR had fallen into wiping each other out, what would 2013 be like? Roughly, we're oversampling the no-nukes outcome by the ratio of our current population to the population there would have been in a yes-nukes outcome, and the less lopsided that ratio is the more evidence that MAD did work after all.
[1] For this wikipedia cites: The global health effects of nuclear war (1982), Long-term worldwide effects of multiple nuclear-weapons detonations (1975). Some looking online also turns up an Accelerating Future blog post. I haven't read them thoroughly, and I don't know much about the research here.
I also posted this on my blog