No, only idiots say that because Proof of Stake is no more or less "centralized" than Proof of Work, they're consensus mechanisms. Ethereum's Proof of Stake beta chain is deliberately engineered for pooling resources and represents a tiny fraction of expected live staking.
Yes, they are consensus mechanisms, and some consensus mechanisms are more decentralized than others. Rock paper scissors is also a 'consensus mechanism'. The question is how many parties are involved and how much influence they have. PoS has fewer participants and has more concentration of influence compared to PoW. In PoS, the stakers are the replacement for the miners who are also the whales whereas in Bitcoin they are much more separated (i.e. miners have a much, much smaller share of the outstanding coins). This greatly increases their influence. (hell, in ETH the developers are are the whales and miners and devs all-in-one) Staking is centralizing not only because all new coins go to the already largest coin holders but also because all but the largest whales must use pools to stake which means they do not have control over their funds while doing this (not your keys etc). If a BTC mining pool went rogue, the miners could switch to another pool in minutes to fight the rogue pool they were just on as their hardware is unseizable by the pool operator. PoS also gives exchanges enormous power. (I hope you trust Binance) PoS blockchains have been successfully attacked by leveraging exchange held coins:
The Steem Takeover and the Coming Proof-of-Stake Crisis
In the cases mentioned in my linked tweets, the pools that blew up were liquid pools, which are an attractive option that allows you to sort of get around the lockup limitation by using another token. (yet another token. tokens to get around the hassle of your other token) If too many stakeholders opt out of staking in a PoS system, because of counterparty risk or fund lock-up or whatever, then the security falls apart, so holders are almost forced to stake. Of course users won't realize this and therein lies another danger.
These guys pooled mainnet ETH on the beacon chain and lost real money. This wasn't test net ETH they got for free from a faucet. They were liquid pools which are smaller and more experiemental but were still mainnet pools. This counterparty risk is simply absent from PoW mining.
To be fair, decentralized staking pools will probably happen one day, just as decentralized mining pools are on the horizon. Also PoS can be 'good enough' security for DeFi chains IMO so long as Bitcoin is there to be the final safe haven long-term store of value destination, so I do own some PoS coins. I'm not suggesting ETH is guaranteed to fail.
