Given the history of Ethereum, what could go wrong with a movement to proof of stake?
https://notes.ethereum.org/@djrtwo/risks-of-lsd
https://research.paradigm.xyz/staking
Another area that is new and can be biased is RANDAO:
https://eth2book.info/altair/part2/building_blocks/randomnes...
There are probably going to be several issues that will arise during and after merge, and might require engineering solutions.
Have a look here: https://twitter.com/web3isgreat