South Africa is an interesting case-rand inflation and a large unbanked population create genuine pull demand, not just merchant novelty. When adoption compounds in that kind of environment it tends to be stickier. Would love to see Lightning vs. on-chain split in the transaction data.