In his post “The risks of LSD”, Danny Ryan identifies control of the NO set (who goes in, who goes out) as the biggest risk from large liquid staking protocols like Lido. After discarding the two most popular options today, changing the set via governance (Lido, Alluvial) or economic selection (Rocketpool), he looks at a third option: a dual-governance, where stETH holders can veto proposals and backstop the safety of the protocol.

In this chapter, he argues that bringing stakers into governance would not add meaningful security because there are many more users of Ethereum than stakers, and they are not the same group. In other words, the stakers don’t represent Ethereum or its users, and hence are not sufficiently incentivized to protect them.

This is a bit of a red herring, and I will explain why.

First, the observation “users and block producers are not the same group” is completely orthogonal to whether stake is delegated or not. It’s simply a property of PoS and all other known Sybil resistance mechanisms (except maybe proof of humanity which is unworkable for other reasons).

Second, there are many ways in which block producers are incentivized to act in the best interest of users. These include

• Block producers want to maximize their revenue; This means maximizing the value of ETH and the demand to build/transact on Ethereum.
• Users constrain the power of block producers by verifying the state (i.e. running full nodes). Censorship can be further prevented by secondary transaction channels such as crLists. Centralized block production, trustless block validation, and censorship resistance techniques are the natural endgame of Ethereum and all other blockchains anyway, as argued by Vitalik in his post “Endgame”.
• It is well understood that block producers don’t control blockchains but are performing a valuable service in exchange for money. Whenever users and block producers disagreed on something in the past, users have always come out on top. The users give a blockchain value and financial activity and hence control the money faucet that all block producers drink from. Aside from taking their business elsewhere, users also have the ultimate weapon of social intervention. PoS makes block producers more powerful compared to PoW but this is a fundamental tradeoff of PoS.

So not only is Danny wrong - if anything it can be argued that PoS creates a bigger overlap between users and block producers than PoW does, and LSD-PoS (where the stake itself can circulate as money/collateral in the economy) creates an even bigger overlap between users and block producers than pure-PoS does. So stakers, and especially LSD-stakers, are very aligned with the security of Ethereum.

TLDR: “Stakers don’t represent users” is a true observation but one that exists in PoS and all other consensus systems. LSD-PoS is not to blame and in fact makes the problem better, not worse.