Data shows monthly L2 dev ranking shifts; top Layer 2 projects by development, Vitalik Buterin L2 critique, StarkNet development activity; linked to Vitalik

Top Layer 2 projects by development: StarkNet leads last month

Based on data from Blockchain.news, which cites a March 2026 development-activity snapshot, StarkNet led Ethereum  ETH -3.74% Layer 2s by recent GitHub-based developer activity. The same report noted SKALE rising, while Arbitrum  ARB -1.09% and zkSync showed slight month-over-month declines.

Development-activity rankings typically reflect engineering velocity and contributor intensity rather than usage or total value locked. As such, they signal where teams are shipping code and hardening infrastructure, not necessarily where users transact most.

Why this matters for Ethereum L2s and immediate implications

As Ethereum Layer 1 scales and fee pressures ease, the case for L2s as pure throughput extensions is being reassessed. At the same time, decentralization gaps, such as reliance on centralized sequencers or multisig-controlled bridges, remain in focus for risk and policy reasons.

The near-term implication is strategic: L2s may need to differentiate on trust assumptions, privacy, latency, or domain-specific functionality, rather than only competing on transactions per second. That shift could reframe how builders choose execution environments and how institutions evaluate operational and counterparty risk.

Editorially, this debate centers on whether the rollup‑centric narrative remains the default or becomes one of several L2 roles. “Rollups as branded shards on Ethereum no longer makes sense,” said Vitalik Buterin, Ethereum co‑founder, as reported by Cointelegraph. His critique pairs Ethereum’s improving L1 capacity with a call for L2s to progress decentralization and specialization.

As reported by Forklog, Arbitrum co‑founder Steven Goldfeder reiterated scaling as core to Arbitrum’s mission while emphasizing settlement security on Ethereum. In the same coverage, Optimism’s Karl Floersch backed faster progress toward higher decentralization stages and a modular stack that clarifies performance–security trade‑offs.

Metrics, Vitalik Buterin’s critique, and L2 differentiation paths

How development activity is measured and interpreted

Most development‑activity indices aggregate open-source signals such as code commits, pull requests, issues, and contributor counts across relevant repositories. These metrics help identify engineering momentum, roadmap execution, and maintenance cadence.

However, they can overweight refactors, infrastructure tooling, or multi‑repo pushes, and they do not measure economic security, censorship resistance, or user retention. Interpreting ranks alongside decentralization milestones, bridge architectures, and sequencer designs provides a fuller risk‑adjusted picture.

What Vitalik Buterin’s critique implies for L2 roles and strategy

In practice, the critique pushes L2s toward clearer roles: privacy‑preserving virtual machines, ultra‑low‑latency sequencing for real‑time apps, or domain‑specific rails for social, identity, or AI‑assisted workloads. It also raises the bar on moving from Stage 1 to higher decentralization, including fault‑proof maturity and minimized privileged controls.

Stakely.io highlights Aztec Network’s differentiation via a permissionless “Continuous Clearing Auction” distribution and a testnet with active sequencer and prover participation. That model aligns with privacy‑first positioning while attempting open, transparent access to network roles.

AInvest notes that institutions are attentive to L2s combining performance with adoption and clear security guarantees. Projects like StarkNet and zkSync are cited for momentum in compute‑heavy or privacy‑sensitive use cases, which may complement Ethereum’s own scaling trajectory.

At the time of this writing, contextual market gauges for Arbitrum (ARB) show a price near 0.1025 with very high 30‑day volatility around 10.01%. Short‑term indicators include a neutral RSI (14) near 41.66, 12 green days out of 30, and SMA50/SMA200 at 0.1369/0.2594 respectively, consistent with a bearish trend designation.