... But I oppose a preemptive PoW change unless there's some sort of new long-term solution.
How about start by prepare guide/reference client in-case hard-fork is needed? Reflecting from Bitcoin-Qt 0.8.0 accident (which is accidental hard-fork due to different DB version), that would help community/developer where to start solve the problem.
I absolutely support this argument.
Although rationally calculated self-interests of players is a vital theoretical support for security of bitcoin, we should be aware of complexities involved.
Things change so fast and, as the paper correctly has established it, adversarial incentives are getting stronger while centralization of mining is reducing the collusion costs to dangerous thresholds. SO, WE SHOULD BE PREPARED.
I have this idea for a long time, a 2-way concurrent PoW algorithm with a 2-3 years smooth migration from ASICs to the alternate cpu/gpu PoW method e.g. ProgPoW. I'm thinking of starting with a 10:1 ratio in favor of sha2 ASICs and a gradual transition to 1:10 ratio against them.
... the hardest part is making sure existing miners won't do anything crazy such as boycott or attack network knowing MAD situation.
Miners would go after their profits if bitcoin mining scene was decentralized and it is possible to improve this factor significantly. We need ASICS for at least next 2-3 years and I don't think we are in a deep antagonistic contradiction with miners, it is rather about pools and
pools, my friend, should be extinguished..
My suggestion is a double attack simultaneously being triggered against pools and ASICs but with different approaches regarding each of them:
Pools should be eliminated from the very first day of the fork but ASICs should be retired with a very small sleepness.
Elimination of pools, sets a huge amount of mining power free and distributed and ready to choose their profits against politics, they should not be slaughtered by a preemptive anti-ASIC fork, instead I suggest a long term mutual co-existence of both algorithms using a double algorithm/double difficulty system which in long term is biased against ASICs, strictly defined by a constitutional rule.