Whilst not quite UPI, the UK has a pay-via-text (sms or Whatsapp) system that nearly every person has access to but barely anyone uses. It amazes me how people naturally gravitate to card-based payments, even when Visa and MasterCard are taking a cut.
Having done a UPI integration for a well-known multinational tech company, I consider it a mixed bag. For the end-user it's a great concept and continues to revolutionize payments in India. It's convenient and it will help India transition away from a cash economy. UPI is clearly the future.
However, UPI currently has a lot of problems. Number one is that to use UPI as a payment method for your product you have to integrate with one of India's banks, which in my experience are almost uniformly technically very bad. They provide almost no documentation, have frequent and often severe service outages, and don't even come close to meeting their availability SLAs. I'm sure over time it will improve but as it stands many of India's banks are not technically capable enough to support UPI with the level of reliability that most businesses require.
Bitcoin is actually more expensive than SEPA transfers or girocard payments, and slower to resolve (and just as impossible to reverse as SEPA transfers).
There is a central provider called NPCI (National Payments Corporation of India). They run the entire network and control who gets on or not. Banks are used as the integration point because UPI necessitated a bank account to run out of. Since you’re directly debiting money out of someone’s account you can’t exactly turn away the bank and when they are on-board it makes sense to use them as an integration partner for legal reasons (as well as customer support).
That UPI was agreed upon by enough people was a happy accident for India.