I find it interesting that the Rest of World article did not mention Lightning network at all, and the working paper for the research referenced it just twice. The survey apparently did not distinguish between on-chain and Lightning payments at all, so there's no data saying that of the retail payments, how much was on-chain versus Lighting network, and as far as merchants accepting bitcoin, how much of that is on-chain online, Lightning network only, or both on-chain and Lightining. To be fair, there was Chivo wallet only added LN support in December, just a matter of weeks before the survey began.

From the working paper:

The survey was done at households in El Salvador. That means no Salvadoreans in the diaspora outside the country were interviewed, yet they are eligible to download and use Chivo wallet. So the survey missed a chunk of the Chivo user base.

Also, the survey only asked about use of Chivo wallet. If you were surveyed, and used Bitcoin Beach wallet, or BlueWallet, or Strike even, whatever, and used it daily, but never used Chivo, you would fall under the "Didn't download" category, which gives unfair weight to the "Chivo failed beyond the free money" and "Bitcoin fails as a medium of exchange" conclusion that was made. Salvadorean minors and foreigners in El Salvador wouldn't be Chivo users either (only Salvadoreans can use Chivo, and even then only those 18 or older).

The survey was at the beginning of 2022, so things may have changed between then and now (May 2022), but:

That means half the surveyed users would not be willing to pay bitcoin's typical on-chain fees (which will always exceed $0.05). Good thing Lightning network exists, so they can use bitcoin essentially free of fees. (Though, because Chivo is a custodial wallet, sending bitcoin person-to-person incurs no fee and payment in bitcoin to merchants who use Chivo also incurs no fee. Withdrawing cash at an ATM, using bitcoin, incurs no fee.)

The only thing I cannot understand is, .... why would there be the expectation that Chivo users would have bitcoin to spend, beyond their initial $30 worth that was airdropped? They cannot buy bitcoin through Chivo, as far as I can tell (I could be mistaken, but I don't see it). And the survey showed that 88% of merchants who had any bitcoin revenues are cashing out into dollars the bitcoin they receive for payment. Nobody is going to go use Strike, for example, to acquire bitcoin that they then transfer to a Chivo wallet, only to then spend bitcoin at a merchant who uses Chivo. Why not just spend from Strike!

About the only use I think I myself would find with Chivo if I lived in El Salvador is that the ATMs can work well as a cash-in / cash-out (CICO) method. With 60% of commerce occurring in cash, adding Chivo gives a way to carry around less cash, but have access to more cash when needed. Similar to how mobile money is used in some African countries, except automated with ATMs instead of using an agent CICO network.