Is Decentralization merely a means to an end? And how does Ethereum address this?

Is Decentralization merely a buzz word and a means to an end?

Ultimately, everyone wants:

1. Availability and fault tolerance: The blockchain continues running even when some nodes are unavailable.
2. Safety: Network operators can't seize or use assets without permission.
3. Censorship-resistance: Network operators can't censor valid transactions.
4. Transparency/auditing: All transactions are recorded on a public ledger and verifiable through permissionless nodes.

Full decentralization isn't necessary to obtain these security properties.

There is no blockchain that exists today with full decentralization. Fortunately, there are ways to acquire security through protocol design without needing full decentralization. Features like L2 exit hatches and recovery through a proof system provide anti-censorship properties regardless of whether the L2 network is centralized.

Other designs like Attestor-Proposer-Builder Separation create a separation of powers where the builder can be centralized, but overall block production is decentralized due to proposers and attestors being decentralized. (This is the opposite of proof-of-work where a miner is single-handedly responsible for building, validating, and proposing the block.)

If decentralization were absolutely important, we wouldn't be using mining pools, staking pools, CEXs, L2 networks, and wallets that rely on RPCs. The Nakamoto Coefficient measures the number of bad entities it takes to compromise a network. If decentralization were absolutely important, we'd be avoiding networks with low Nakamoto Coefficients like Bitcoin and Ethereum. And we'd be avoiding networks with low client diversity, or with high miner/validator hardware requirements.

How does Ethereum addresses security?

Despite having a low Nakamoto Coefficient, Ethereum fundamentally still has all the benefits of decentralization because it is designed with:

• Very high economic security and safety from Proof-of-stake. Aligned economic interests where the people holding the token are incentivized to protect it. Cannot be attacked by miners without economic stake.
• Very high client diversity with 10+ independent client dev teams to allow for fault tolerance and high decentralized governance
• Deterring malicious actors by slashing validators who act badly. This increases economic security.
• Providing Economic finality through its Casper-CBC finality gadget, which also prevents long-range attacks and deep reorgs
• Providing sustainable security budget through EIP-1559 and validator rewards to support network security indefinitely
• Attestor-Proposer-Builder Separation: Ethereum builders are centralized. However, that doesn't matter since there is attestor-proposer-builder separation where the builders are separated from the nodes proposing the blocks, which are also different from the nodes validating the blocks. This separation will be further enshrined in the upcoming Glamsterdam update.
• Layer 2 rollups at Stage 1-2 inherit the security of L1 despite having centralized sequencers because honest transactions can be forced, and dishonest transactions can be economically-reversed through proof mechanisms even without the support of the centralized sequencer.

How does Bitcoin addresses security?

And despite having a low Nakamoto Coefficient, Bitcoin has ... umm ... none of these.... Sorry, never mind.

I'm sure the adoption of Stratum v2 miner-submitted block templates will eventually pick up by the year 2050.

And we can only hope they figure out a sustainable security budget protocol.