Asicboost is a successful attack on the Cryptographic POW function of bitcoin because it decreases the time complexity of the work. To say its an engineering optimization is a lie… In cryptography any algorithm and reduces the time complexity of solving a cryptographic problem is considered a successful attack.
It is very dangerous to leave this bug in the protocol, who knows what kind of economic effects it will have in the future. What if it becomes beneficial to artificially construct transactions so there are more collisions to reuse in asic boost….
The only way to block both ways of asic boost is via a Hard fork. We may as well increase the blocksize to 2mb in this hardfork. Then we can safely soft-fork to segwit. This upgraded is as per the hong kong agreement, which majority of the miners signed.
If they appose the 2mbAsicBoostBlock then its more then obvious they are using the covert asic boost on the main net. As Andres said when mining profit margins are less than 1% the advantage asicboost gives is 20 fold. This is a serious threat the to free-market forces that protect bitcoin and both large blockers and small blockers will support this plan. Anyone who doesn't is exploiting the bug. Is anyone interesting in starting a BIP to implement this?