Ethereum News Today: Ethereum L2 Debate Intensifies as Projects Shift to External Data Availability Models

The debate surrounding the classification of Ethereum layer-2 (L2) solutions has intensified as projects built on the OP Stack introduce alternative data availability (DA) models. These so-called “Optimiums” finalize transactions on Ethereum but source their data availability from external networks such as EigenDA, Celestia, or Avail. This shift raises questions about whether these chains should still be considered traditional L2s, given the additional trust assumptions they introduce [1].

Toghrul Maharramov, a rollup researcher and former Scroll contributor, has taken a firm stance against the expanded definition of L2. In a recent post on X, he declared, “Optimiums are not L2s. Anyone who says otherwise is trying to psyop you.” His argument underscores the concern that broadening the L2 label could undermine the core security principles that Ethereum is known for [1]. A notable example is Celo, which transitioned to the OP Stack with EigenDA in March 2025. It is now categorized as an “Other” by L2BEAT, an influential Ethereum scaling tracker. This classification is reserved for chains that fail to meet strict rollup criteria, such as having fewer than five external actors who can submit fraud proofs and lacking a DA bridge that posts availability proofs on Ethereum [1].

Lisk, which migrated to Ethereum in late 2024, faces similar scrutiny for its reliance on a permissioned challenger set. While it does post data on Ethereum, it does not meet the same trustless standards as traditional rollups. Sreeram Kannan, founder of EigenLayer, explains that rollups provide “out-of-protocol scaling for execution,” while EigenDA offers “out-of-protocol scaling for blob space.” Both exist outside Ethereum’s protocol, but only rollups allow for direct on-chain enforcement of validity [1].

According to L2BEAT’s current methodology, an L2 must allow Ethereum to independently verify data availability and reject invalid state roots. Optimiums that lack these properties remain in the “Others” category. However, some may eventually graduate to a more secure classification—such as “Validiums & Optimiums”—if they introduce robust fallback mechanisms or ensure that data availability failures can be objectively attributed [1].

The debate extends beyond technical definitions to real-world implications for users. Kannan acknowledges that while critics argue L2s are controlled by multisigs, similar concerns apply to applications on other blockchains like Solana. The broader challenge lies in minimizing trust across the ecosystem. Until reliable solutions emerge, the conversation will continue to focus on trust models and their impact on fund security [1].

Source: [1] Optimiums and the 'L2': Ethereum security revisited (https://blockworks.co/news/optimiums-and-the-l2-ethereum-security-revisited)