Here's an articulate presentation on these qualities of Bitcoin that many argue are it's primary function, viewing it as an ideal store of value[0].
Here also is a book by George Gilder on the flaws of fiat currency vs sound money[1], but which also covers a much broader spectrum of economic issues we've face in recent decades, and highlights qualities of Bitcoin (and to a lesser extent gold) that counteract these trends.
They both support the argument that Bitcoin is not primarily useful as a high volume, high speed payment network (though 2nd layer solutions can serve this purpose, as well as a decrease in the rate of price appreciation long term), but rather as a store of value, mainly concerned with locking value away from central banks and governments, and in practice would as you highlight be most useful for higher value lower speed transactions.
The economic argument of sound money is that it promotes low time preference [2] vs inflationary money incentivizing high time preference.
[0] https://youtu.be/ukjDCeuNK3Y
[1] "The Scandal of Money - Why Wall Street Recovers but the Economy Never Does"
https://www.amazon.com/Scandal-Money-Street-Recovers-Economy...
and why do we care about those?
Finite money supply, I get. I'm not I agree but I get it.
Why exactly do I need an immutable record that I paid some guy 10 bucks for a souvenir shirt at a concert? That I stopped and grabbed a coke at mcdonalds on the way home. That I went and bought some legal weed.
Small-value and/or micro transactions are arguably never going to work at least in the way most people think about them, though lightning network may succeed in this area, but it's not the most important aspect. Instead, the existence of a money whose supply cannot be increased at the whim of political expediency is the true value.
Maybe you already know this, but there is a strong debate amongst the Bitcoin "community" of whether to accept higher fees, less transaction throughput, but greater decentralization and security or to worry a lot about fees which are rising, and increase the block size "now" because a lot of these people are upset about fees being too high for small value transactions such as the ones you seem skeptical about.
They would argue the tradeoffs aren't that bad, and also that scaling on chain can be achieved to allow for high throughput and low fees. Many don't think this is possible to even approach the volume of something like Visa so why bother and besides 10 minute block times make retail purchases challenging anyways.
Immutability comes with the territory of irreversible and unforgeable. Why you need immutability gets into the argument of why proof of work is the only way this system can work. The double spend problem, consensus, the decentralized blockchain etc etc which I'm sure you're familiar with but anyways
I encourage you to read the Gilder book, it's not that long and he's well known for seeing innovation trends before many... the value of sound money is pretty profound in my opinion. The fact that not only is Bitcoin scarce but completely explicit in its monetary policy (though there will be lost coins that fudge this a little) provides the best form of money as measuring stick that we've ever had.
TL/DR* You don't for small purchases, that use case is already solved today. Finite money supply serving as superior store of value is the point.
You think venezuelans should be forced to use their local currency?
You think Zimbabwe's people should go thru hyper inflation again? They are about to, and bitcoin is booming there as an alternative.
Sorry, you seem to think you know better than other people what they should value and what they should do.
Bitcoin is completely voluntary, you don't like it, don't participate.
It's also not your energy getting spent.
At the risk of sounding like a libertarian: Other people are not your property.
It isn't my energy being spent, but it is CO2 that harms our planet. Until we have zero-emissions energy I think it is reasonable to criticize systems as energy-inefficient.
I'd add that I hope we will eventually tackle the energy problem, hopefully without throwing away the good parts of decentralization / distributed ledgers.
It might be optimistic, but I think many early technological revolutions bring learning costs and risk environmental damage.
The benchmark for Bitcoin are normal financial transactions as they exist today (and work quite well, most of the time). So the question really is: Are you ready to have humanity spend the additional energy consumption of a country (or many countries if Bitcoin were to scale up) to run a system that offers only very limited throughput (relative to established systems) and hasn't even proven yet that it's even a viable alternative to fiat money from an economic perspective? Not to mention the scaling issues that come with such broadcast protocol.
I don't really see the widespread use, at least as envisioned by Bitcoin proponents. If it's just the decentralized aspect and transparency that you want, you can relax your trust assumptions and implement something like a distributed ledger with known identities which doesn't need proof-of-work to prevent sybil attacks (identities which of course would have to be certified in some way, unfortunate).
Oh well, it's still a very cool technology, but I hope for the planet's sake that cryptocurrencies stay in their niche of anonymous transactions for those who actually need it.