In the general case I don't think that majority miner attacks can be defended against without changing the hash algorithm.
Specific cases like "empty blocks" could be addressed via a hard fork.
But there's nothing stopping miners from faking transactions in blocks, sending amounts to themselves, filling the block with OP_RETURNs, or just doing the minimum possible to get around your "fix".
PoW absolutely relies on >50% of miners being honest, always has done.
The purpose of this project is to challenge people to be specific about the details of that hard fork. Currently there are no theoretical proposals, or concrete BIPs to address.
It depends how rationally the market responds, I guess.
A reorg attack and a DoS are equivalent in my mind - anyone who's capable of pulling off a lengthy DoS is also capable of performing a reorg attack, whether they actually do or not is kind of irrelevant.
there is a type of reorg attack I proposed that's now been dubbed a "purge attack", that has the effect of destabalising the network by creating a window for all affected users to double spend their previously "finalised" transactions: https://medium.com/deribitofficial/destabilizing-bitcoin-con...
Specific cases like "empty blocks" could be addressed via a hard fork.
But there's nothing stopping miners from faking transactions in blocks, sending amounts to themselves, filling the block with OP_RETURNs, or just doing the minimum possible to get around your "fix".
PoW absolutely relies on >50% of miners being honest, always has done.