How much USD$ or equivalent is moved into cryptocoins and how much of that amount is considered fraudulent? Until you have numbers there, then I simply don't believe your premise.

There's four actors here. Victim, Criminal, SendBank and ReceiveBank.

Criminal initiates $X on behalf of Victim (or tricks them into doing so) using SendBank. Criminal then gets the money at ReceiveBank and usually withdrawals it before it can go noticed. Criminal therefore commits fraud against victim.

Since both banks are in the interest of keeping Criminals out and Victims (well just normal person/biz), they usually offer the ability to safeguard the Victim against financial harm, usually by paying the Victim money their money back and eating the cost. Hence why KYC is so important on both sides.

This is all conducted through a trust system that crypto fans love to hate because it's "centralized". AFAIK no crypto coin offers this ability because the trust system is inherently built on the idea that no person would ever be tricked into sending money to someone they didn't intend to, which is, ummm a downright terrible assumption about real world behaviors.

They've tried! ACH has been somewhat modernized over the decades, in the form of compliance and in-protocol changes (faster turnarounds, additional transaction codes for online transactions and different cardholder verification methods). The Fed has also been working on a modern replacement to ACH for at least a decade, and is planning on performing an initial release in 2023[1].

But as for why it's taken so long: banking in the US is, for various reasons, significantly more complicated than banking in most other countries. The US has thousands of FDIC-insured banks and credit unions, each of which operates under a patchwork of municipal, state, and federal regulations. Banks are (mostly) required to honor each other's checks and transactions which means that, in the worst case, there's a total graph of all ~10,000 banks and credit unions serving as both ODFIs and RDFIs[2]. Transaction failures between any two nodes in that graph are a possible compliance failure, so any common mechanism is both a lowest common denominator and resists any modernization or other changes than can cause disruption.

[1]: https://www.federalreserve.gov/paymentsystems/fednow_about.h...

[2]: It's actually simpler than this, since the ACH clearinghouses serve as a central resolution and dispatch service for all ODFIs and RDFIs. But each O/RDFI still has to interpret the ACH record(s) they receive, so there's still extraordinary inertia against any changes.

> Bitcoin would be calculated by value deemed as being defeauded from users onchain as payments and on exchanges, that year, and compared to a portion of its.. marketcap? or by quantity of non-fraudulent transactions?

Yep, this is a serious problem when quantifying fraud in Bitcoin and other cryptocurrencies. It's concerning in its own right that we don't really have a reliable, standard way to quantify fraudulent behavior in cryptocurrencies.

> Since bitcoin is all one kind of payment method (well lightning would be different), it would have to be compared to the other kinds mentioned in the nacha article too, ACH + ATM + Card fraud

This doesn't make sense on two fronts. First, ATM fraud is ACH fraud, since ATM transactions are settled by ACH. In other words, you'd be double-counting there. Second, payment card fraud is diversified: it's either debit card fraud (which is also ACH-settled, and thus would be double-counted) or credit card fraud, which is fraud against the lending party and has its own (even simpler!) dispute resolution process.

All told: we need actual numbers to make a coherent comparison here. In the absence of that comparison, ACH's performance on an absolute scale is admirable.

Or immediately buy something else with those bitcoins, or immediately invest in something else, do both while keeping some for savings etc

Some people will never get it