Perhaps it is impossibly naive to think that centrally controlled bank systems could be so straight-forward and simple, free of precedent and regulation, but I'd like to understand why.
I'm sure we could build a better banking system if we scrapped everything and started over. But what we have is really good enough for most people.
Having worked with the banking industry for over two decades, the banking industry is capable of a lot. They have infinite capital and they hire some of the best people on the planet.
Fundamentally Banking is risk-averse conservative culture. Their primary focus is "good enough for most people" then don't break it. Absent government regulations from a major catastrophe or an existential competitive threat, they will not change.
It's just that there are some minor disadvantages (a bit of profitability back when higher interest rates meant that any delay brings interest revenue to the bank, but mostly customer habits - no single institution wants to be the one to try and change them) so USA banks won't make such a change unless forced to.
This suggests to me that there's a intrinsic cost to supplying the guarantees that a person gets with a wire transfer over something like an ACH. And that's probably why ACH has the shortcomings that it does.
There is some intrinsic cost for a 'true' wire transfer, but (1) it's not due to the benefits/guarantees that the customer gets, but due to legacy/compatibility issues requiring a not-that-scalable process, lack of international cooperation, and the inefficiencies of bilateral settlements (or, often, a chain of bilateral settlements) vs a centralized bulk settlement system; and (2) it's not in line with the prices charged. $30 for a wire transfer is a reasonable market price in many markets, but the fully loaded cost for a wire transfer (assuming a electronically initiated proper payment without any missing information so it gets 'straight-through processing') that I recall is something comparable to $2, the rest is profit - the market for this service is simply not competitive (people choose their banks based on other factors, not the level of rare fees like this) so that's a revenue stream that nobody wants to let go without a good reason.
The key thing for preventing money-forwarding scams, however, is that all the commonly used payment systems in your society must be "reliable". If some are, and some are not, then the intersection of these payment systems opens people up to an "arbitrage" of sending "reliable money" after receiving "unreliable money"; you won't eliminate this type fraud by making a better payment system accessible, that will happen only after systems like cheques are driven out.
With regards to the exact scenario you outline, the core problem is that ACH payments in the US cannot, at the receiving bank, be guaranteed to represent good funds except probabalistically after the ACH return window has passed, and may not be good funds even then. There are both legacy reasons why this is true, dating to the clearinghouse system for paper checks, and UX reasons why it remains true.
Many remediations against this and similar attacks create winners and losers in the legitimate economy. For example, increasing the friction to receive a transfer from a novel source impacts the ability of family members to help relatives in event of emergency, hits the cash flow of semi-formal businesses severely (many sole props operate out of personal accounts and may take dozens or hundreds of first-time payments a month), affects new clients in the onboarding window (who you can’t usefully exclude because they make up a large chunk of accounts which will participate in this fraud), etc.
Note that another way to phrase “Ensure customers don’t end up holding the bag” is “Banks should be less willing to extend credit to poor people” and that there are non-trivial policy and justice implications of that proposal. It is not widely understood by consumers that a US checking account is a credit product, but it is, and the operation of many households assumes this fact of the status quo.