This is a bit of a tangent but you mention that Bitcoin's price would be unaffected. Personally I don't foresee that. It seems to me that when knowing that an attack is underway, some Bitcoin investors would already start to get uneasy, and some would sell their bitcoin. And since this will cause the price of Bitcoin to drop, the 51% attack will then only be easier to afford in theory, as the mining rewards would be decreased and some of the honest miners therefore ought to fall off. This could create a feedback loop where, once the attack is looming in the near future, more and more Bitcoin investors will migrate elsewhere, potentially to Ethereum.
if the network knew the hash power was to get 2x competition(network doubles where half is a colluding group).. meaning older miners get half as much sats... its the same as a halvening event in regards to miners prospects.. we already seen miners re-evaluate global bottom value change from $25k per btc to now being ~$50k bottom due to reward halving this year
so it ends up the same thing. if miners have X cost but getting X/2 rewards then they wont sell the rewards at X they would only sell for x*2 to break even
then new users wanting to get into bitcoin but now unable to mine due to al batches for Y months are sold out. will work out the prospective costs of mining and then just buy coin on the market.. again causing a market rise effect
as we have already discussed the implications of a multi-month ramp up of batches to get a colluding group to 2x the hashrate/difficulty to get 51% of the network.. to match/beat the honest half of the network.. those honest miners would take that opportunity to buy more coin cheap at the $50k-$70k(current) knowing that the impending hashrate /difficulty growth and reduction in sat rewards would cause the market to INCREASE
and as said whilst the lead up to the collusive gangs trigger date for the attack, those getting early batches would honest mine and try to get ROI and end up not wanting to shoot self in the foot at trigger date because now they are highly invested in bitcoin and wont want to lose their investment
(in short a influencer cant convince hundreds of random ethereum users to collude becasue users end up caring more about their own investment, instead it would only work if there was one mass eth holder that doesnt care about losing investment because all he cares about is 'taking bitcoin down')
anyway, back to discussing a scenario of an impending attack in the next X months:
in that same time as the eth stakerconverting to be bitcoin miners.. exchanges would look into solidifying their deposit algo's to not allow new deposits to trade unless it meets X confirms. and then even when trading on exchange mysql databse balance. not allow withdrawals for X time to ensure that any nafarious pool trying to re-org the blockchain would need to re-org many many many blocks back.. meaning (as mentioned before) the colluding group would need far more then 51% of the network to be able to re-write blocks very far back
if you run the math of average block speed and how far back a 1% speed benefit means in reality to going backwards and then moving forwards to catch up and over take.. 1% benefit over honest network wont get the colluding group to go that far back. so services can just enforce the "6confirm rule" or a 72 hour wire transfer delay(if aim was to sell coin for dollar and cash out) to evade colluding pools from trying to double spend
You seem to be of the opinion that the price of Bitcoin is directly dependent on the cost of mining. By that logic, the Bitcoin stakeholders could always just spend a small percentage of their wealth at any given time to increase the combined hash rate of the world, and then see their assets grow by a similar factor. You do see that this logic is flawed, don't you?
I've already addressed the point that the miners loss of CapEx in terms of not being able to sell their ASICs after the attack has to be compensated by the stakeholders, and that this is part of price of the attack.
In terms of Bitcoin extending their confirmation period, you bring up a valid point. Bitcoin would indeed likely extend this period if an attack is anticipated.
However, they cannot extend this period
retrospectively. And because the operational costs of an attack are relatively small compared to the full price (and not least compared to the potential winnings of a 'Rival Goldfinger attack'), the attackers could in theory rewrite
months of the ledger in an attack. So in order to mitigate an attack by extending the confirmation period, Bitcoin would thus have to extend it to months. And if Bitcoin does this, its use as a cryptocurrency will then be close to none, which would therefore cause severe damage to its value already.