Rolling with the mine analogy, a 51% attack is equivalent our group that sent in a million miners changing the diamond mine into a 'glass' one. It becomes fake, and loses nearly all of its value and people move to the new 'real' mine. You do nothing except critically hurt yourself.
Forks don't help with 51% attacks as the attacker can continue to use their hardware on the fork. Further, you can use a proxy to hid the origin of a block, so Bitcoin would need to move to a new hash which would take a long time and prevent any obvious successor.
There are numerous different hashes with varying levels of ASIC 'resistance' - yeah it's a cat and mouse game, but hardly a major issue. The 'obvious' successor would be a matter of the unpredictable public as I'm certain numerous entities would vie to become e.g. Bitcoin 2.0. In the end the market would decide which was the winner. The great thing about it all being that the users could actually come out net winners in the end as the successors would likely not only be technical superior, but it would also be able to use the exact same ledger (as with Bitcoin cash for instance) so that you start with the same relative share in it as you did with the original.
In a way it would even be a good thing as Bitcoin itself is increasingly dated technologically, but is dominant because of its market positioning - which makes it difficult to change. If its market positioning was damaged, we could see the rise of an improved successor.
The tiny block size is arguably an ongoing 51% attack from an oligarchy that is in collusion. Again, depending on what you consider centralization having a tiny group that's in direct communication easily qualifies.
Pas: In your example if the only option is to use picks from one company or your hands because other tools are confiscated at the enterance then it's very centralized.