Below is an analysis of the decision-making process:
1) If the small block of bitcoin can guarantee decentralization of bitcoin (though it is still uncertain whether decentralization can be achieved), bitcoin can survive despite of the governments hostility or not. However, the small block will restrict further development of bitcoin.
2) A large block will expose bitcoin to risks of losing its decentralization. If the government is against bitcoin, bitcoin is doomed to die.
3) If bitcoin is expanded into a large block and the government is not against its existence, bitcoin will fare better and its development will be unlimited.
http://i.imgur.com/OfXRbWL.png(Whether the big block will result in centralization is not discussed in this article. This article is written in the mind-set of opponents.) Because we cannot know whether the government is against bitcoin, the safest way is to maintain decentralization of bitcoin as much as possible no matter whether its development is the best. This is a typical prisoners dilemma.
That's not Prisoner's dilemma, what you got there (from looking at the picture) is more like Pascal's wager. The decision matrix looks like this:
God exists (G) God does not exist (¬G)
Belief (B) +∞ (infinite gain) −1 (finite loss)
Disbelief (¬B) −∞ (infinite loss) +1 (finite gain)
Pascal's wager, like your proposition, is misleading.
Your variable valuation is arbitrary: big blocks don't lead to centralization, assumption that non-mining nodes contribute to decentralization is unjustified (and unreasonable), as is your assumption that non-mining nodes would be the point of attack, etc., etc.
Garbage in, garbage out.
