Why is a validator punished for downtime in Proof of Stake?

So in ETH PoS scheme (and in the PoS scheme of other cryptocurrencies), staked tokens can be slashed if the validator is offline for too long. But why? Why can't the network be designed to be resilient to such failure? In Proof-of-Work, no miners or full nodes get punished if they just stop. They only stop receiving rewards when they stop, not getting their existing profits retracted.

I'm concerned about this because this "always-online" requirement scares away most people from running a validator on their own (that plus the 32 ETH is too expensive). This encourages centralization.

Official docs:

For example, a user can lose a portion of their stake for things like going offline (failing to validate) or their entire stake for deliberate collusion.