Something I don't think is thought about enough in the world of stablecoin-based cross-border payments are shifts in trust assumptions. For example, a cross-border payment provider operating on stablecoin rails who seeks to abstract away the user experience to "it is significantly faster and cheaper to transfer value across borders with different currencies" still has to deal with the fiat on/off-ramp problem. While a pure P2P-matching system (the platform matches counterparties in the receiving jurisdiction) keeps the trust assumption with the platform, it is often the case that the platform assumes a quasi market-making role, resulting in them holding local fiat funds in local banks. The more they seek to round off jagged P2P match-making, the more reliant they become on local banking infrastructure. This is mechanically quite interesting...