ChainFeeds PRO Covers Builder Economics, EIP-7999 and Native Ethereum Delegation

ChainFeeds Research’s PRO #155 gathers Bitcoin protocol updates, Ethereum governance research, fresh papers, and MEV-related work in one issue. On the Ethereum block-building side, Christoph Rosenmayr and Jascha Samadi look at the supply side of the market and ask a different question: not who built the most blocks, but who captured the most valuable orderflow. Using Dune data, they say validators have earned about $1.36 billion in execution-layer rewards through MEV-Boost since early 2024, while builders kept about $404 million in builder surplus. That surplus is not spread evenly. Titan, the report says, currently builds about half of all blocks but captures roughly 80%–85% of total builder surplus, with a surplus-to-validator-reward ratio above 80%, far higher than the roughly 20% seen across most competitors. The paper frames the real scarce asset as orderflow that carries high priority fees, strict ordering demand, and MEV value. Builder competition, in that view, is a two-sided service relationship with searchers, wallets, trading bots, solvers, and other orderflow providers. The issue says large orderflow sources often route transactions to a small set of builders in exchange for lower information leakage, accurate ordering, conditional execution, low latency, privacy, and priority-fee rebates. It also notes that builders may subsidize blocks by paying validators more than the block is worth, so more blocks do not automatically mean more profit. Priority fees are heavily concentrated at the front of blocks: the first 5% of transactions account for roughly 35%–65% of priority fees. In Bitcoin research, Shinobi discusses how users could recover BTC locked in hash-addresses if quantum computing breaks today’s ECDSA signatures, while avoiding protocol-level confiscation of coins that have not moved for years. The suggested approach combines several mechanisms so that any one proof path can authorize a spend. These include using BIP32 xpub derivation relationships, pre-signing a migration to a quantum-safe public key with an on-chain timestamp, and a commit-reveal migration flow. The proposal assumes the public key or Taproot internal key remains hidden. Because many lightweight wallets leak xpubs to third parties for balance lookups, Shinobi suggests new derivation paths and per-address Electrum-style queries instead. Bitcoin Optech Newsletter #418 also covers Antoine Riard’s CMTC, or Conditional Message Transfer Contract, a proposal aimed at Lightning Network channel-blocking attacks. If an attacker opens HTLCs or PTLCs and then stalls, they can occupy payment-path liquidity for free. CMTC makes the withhold fee rise as the delay grows, using block height as the timing reference. The contract handles normal delivery, receiver offline cases, and situations where the sender fails to deliver the message. The same newsletter notes that Bitcoin Core has started shipping statically linked test binaries for bitcoind and command-line tools on x86_64 and aarch64 Linux. Unlike dynamically linked builds, the static packages bundle their dependencies and do not rely on the system’s glibc or other shared libraries, which should make them more portable across older Linux distributions, Alpine Linux, and minimal containers. Bitcoin Core also merged a relay-throttling change that replaces per-peer transaction announcement queues with a global token bucket. The old design could trigger repeated re-sorting across many peer-specific queues when transactions arrived in bursts, creating a DoS risk. The new design keeps one shared backlog queue, sorted by fee rate, and then feeds a smaller peer queue for privacy batching. On Ethereum, RIG researcher Anders Elowsson examines EIP-7999 and the compatibility problems that come with multidimensional fees. Future transactions may need to pay separate base fees for execution, state creation, and data, but the EVM and many existing contracts still assume a single scalar gas budget. He compares four paths: Aggregate EVM gas, Multidimensional subfee market, Universal overflow, and an Updated EVM. His focus is not choosing a single winner, but separating the trade-offs around scalar introspection, callee gas caps, post-call headroom, and who bears the risk when resource prices diverge sharply. Jeff’s Native Ethereum Delegation (NED) proposal starts from the concentration created by today’s staking stack. Users who choose an exchange, an LST, or a staking service often also decide which validators their ETH ultimately reaches. NED would move delegation into a protocol-native public pool: users would delegate through NED without naming operators or validators, and the protocol would distribute principal across willing validators. Commercial providers could still compete on custody, liquidity, insurance, and UX, but would no longer be the direct destination for consensus weight. To limit identity splitting, NED uses a linear allocation rule, Di=uBi, so splitting one balance across multiple validators does not change the total assignment. That alone does not solve incomplete participation coverage, so Jeff adds the Delegation Concentration Envelope, or DCE, a worst-case bound on how much delegation can be packed into any eligible stake interval. The design does not try to identify real controllers. It instead caps the worst-case distribution and keeps hidden coalitions from crossing the safety threshold merely because of NED routing. The paper presents NED as a research concept, not an EIP. The issue also highlights ragged multi-instance GKR for Poseidon2b. The idea is to avoid either running separate proofs for each region or padding everything to the maximum width. By using an implicit selector for missing high-order coordinates, the prover can keep one unified GKR walk and avoid physical padding. In experiments across nine Poseidon2b regions of different widths, the ragged setup cut median prover time from 17.609 seconds to 10.830 seconds, reduced verification time from 4.896 seconds to 0.802 seconds, and shrank the algebraic transcript from about 2.27MB to about 363KB. The MEV section points to Flashbots’ new newsletter and several related items, including a paper on the geography of block building, a study on collective problem decomposition, a proposal on batched-threshold encrypted mempools, two Arbitrum bot-analysis pieces, a fee-market analysis for Glamsterdam, and talks from the MEV-SBC ’26 workshop, BASS Stanford, and Encrypt The Mempool #7. The papers section closes with “A contribution to the critique of blockchain censorship,” a paper from Derivation Technology Ltd, Aarhus University, and Nanyang Technological University. The authors argue that once an attack-detectability threshold is added, whale validators may get lower returns from joining a censorship attack, changing the equilibrium and making it harder for small validators alone to launch one. Monte Carlo simulations suggest the mechanism could reduce the incentive for whales to participate in censorship on both Ethereum and Solana.