You can safely store data using any number of technologies, e.g. the traditional data stores that hold bank accounts and other critical financial information. Every day, trillions of dollars in wealth are moved around using these systems, and they run the everyday lives of basically everybody on the planet.

Blockchain doesn't guarantee a server on the network can't change anything--it absolutely can--it merely creates a consensus among nodes that you do not trust.

In other words, if somebody was able to simultaneously hack most of the Bitcoin nodes all at once, they would effectively control all of Bitcoin. The reason Bitcoin (and other broadly used cryptos) are safe from this attack is that they have so many nodes that such an attack is extremely unlikely.

This is what blockchain is for. Now you can understand how ridiculous that is to use for two known and trusted entities to communicate with each other.

As for whether a government could implement some sort of solution for very low volume transactions, well, I guess if you throw enough money at a problem you could do almost anything. But I suspect the people working on the project will wonder why they are wasting their time like this. And the project would get bogged down when they discover that their project has all of the other safety and security challenges of a financial project like this in addition to dealing with blockchain.

For me, the CBDC silence right now is deafening: no major country has achieved any sort of ubiquity with their CBDC, they are all currently in the exploratory phase, or in a pilot phase, or adopted very sparsely.

I predict there will not be a mainstream CBDC until countries figure out they need to drop blockchain.