These relayer design tradeoffs are important to understand as Cosmos expands IBC to non-Tendermint chains that have longer time to finality (e.g. Ethereum). We need solvers that can handle faster-than-finality transfers, and they need to be able to handle lots of volume Hart's thread buried the lede a bit imo. The breakthrough isn't just bond vs. escrow—it's replacing slow, permissioned oracles with quick, user-driven verification secured by economic incentives Bonding flips the incentives: the relayer puts up collateral rather than the user, so the system can trust user verification instead of waiting an hour for UMA Bonds get recycled instantly → way less capital sits idle → solvers can process more volume Example: Bob bridges Base to Ethereum. Instead of waiting an hour for a verifier, Alice sees Bob's transaction complete, references it in her own, and immediately frees the bond The tradeoff: relayers need enough liquidity on their balance sheets to cover peak traffic, though the design includes just-in-time bonding to help manage this Smart, elegant approach. Nice work @hal2001 & Across team