Formalizing decentralization goals in the context of larger L1 gaslimits

Vitalik Buterin argues that raising the Ethereum L1 gaslimit, and the higher full-node hardware and bandwidth requirements that come with it, is likely a net good, and that Ethereum needs a clearer framework for its decentralization goals before doing so. He proposes three node types: light nodes that verify the chain and are easy enough to include in default wallets, medium nodes with an n/2-of-n trust assumption such as attesters and FOCIL participants, and heavy nodes with a 1-of-n assumption such as block builders and provers. Heavy requirements on block building are expected to be more controversial than those on proving, since a malicious dominant builder can cause harm. He also says bandwidth limits should be more conservative than hardware limits, and that the numbers remain to be filled in.