Before you dismiss Elysium as another L2, you have to understand how the underlying chain works to see the value prop/use cases. HyperEVM produces small blocks every second with a 3 Mgas/s limit. This works relatively well without much load, but easily becomes congested under load. That means higher effective fees, slippage, fill time, etc. To have a chance at becoming competitive to HFT style quoting, you need smaller blocks (100 ms on Elysium) and a higher gas limit (300 Mgas/s). That is what lets market makers refresh quotes 5–10 times per second cheaply instead of being bottlenecked by HyperEVM’s intentional constraints, so ordinary swaps get tighter spreads and more reliable fills. This is also what is needed for more TradFi-like DeFi applications—onchain order books, continuous quote streaming, matching engines, and other high-frequency market infrastructure that simply cannot function under HyperEVM’s 1-second blocks and tight gas limits. Market makers can actually stay competitive with the HyperCore book instead of constantly lagging behind HyperEVM’s 1-second blocks, while still settling back to HyperEVM with HYPE as gas. Finality is softer than native HyperEVM (optimistic rollup with a challenge period), and withdrawals take longer, but for high-frequency quoting and swaps the speed and capacity are the point. The tradeoff is a single sequencer and softer finality, but that’s what buys the 100 ms blocks and 300 Mgas/s needed for this workload. I think the imposed constraints on Hyperevm make an Elysium like L2 necessary.