There's one thing that Bitcoin does better than almost any other cryptocurrency right now, and it's being secure against consensus attacks. Ethereum comes close, Chia arguably surpasses it, but otherwise most cryptocurrencies could very easily succomb to collusion and majority attacks.
Proof of stake chains are virtually immune to collusion and majority attacks. A single actor would need to own huge amount of resources and then essentially burn the coins. To purchase such a dominant stake the price would rise very quickly... making it very unlikely to be feasible. In comparison, attacking the larger Proof of work chains only requires a few $100M in hardware (the harder ones) and a MW-scale power source used sparingly. I don't see why PoW is still used...
They are not immune at all. The security assumptions are different.
> A single actor would need to own huge amount of resources and then essentially burn the coins. To purchase such a dominant stake the price would rise very quickly... making it very unlikely to be feasible.
You are assuming here a single-sided market where the existing majority holders are not interested in realizing profits on their holdings. For many cryptocurrencies this is not the case, especially the hyped VC-funded ones.
If you are a large majority holder, selling off to a known exploiter is a very risky move. You might be able to realize great profits of the price increases. But you don't know how much the attacker has gathered from other sources. I think realistically it can happen, but a takeover attempt from buying a large stake should be plainly obvious. The community could even organize price fixing (i.e. "hold", which we've seen on WSB) to fend off the attack. You can even pull off something like forking the chain to invalidate the funds of the attacker -- it would be pretty obvious which stakeholders are responsible for the attack. For a chain where those coins are tainted or deleted from the chain.
Those avenues are not really available for PoW chains... if an attacker gets the necessary hardware, there's not much you can do. You can try changing protocol, changing algorithms, etc. but this has obvious consequences of its own (making all mining hardware so far near worthless)... it's a bit scary to be honest.
