Recognizing it's merely an example, I think nonetheless it would be good to present examples that are actually compelling.
Let's think about your example and what it entails.
So you're going to a pizza party with your friends, and instead of saying to your buddy Mike who you're about to see in person and share some pizza with, "Hey Mike, here's 5 bucks, thanks!"
Instead you fire up the ol' ethereum whizbang machine, and "pay" Mike. The payment then rattles around the ol' globally distributed whizbang machine, at the SPEED of Light! powered by electricity through countless microscopic electronic gating mechanisms before finally settling into the storage media of thousands of computers around the world, where it will stay, forever.
That seems needlessly wasteful of resources and expensive.
What it boils down to, is what is this good for?
While some may have been, not everyone was confounded by the introduction of the web. Many could see obvious avenues of potential and development.
Even back in the 60s or 70s people were envisioning computer network enabled home shopping. Not sure many were envisioning a global distributed turing complete whatchamacallit with ever expanding resource needs because of no garbage collection by design.
So simply, what's the point? What do you do with it besides buy pizzas with your friends? How is it better than existing methods for doing whatever you envision doing with it?
What if anything, does it enable that would otherwise be impossible?
You could make a similarly incredulous argument that messenger apps are totally ridiculous because why would you use a 500 dollar handheld computer to write a message through the "internet" when you could just talk IRL?
I mean, like, within three years you'll probably be having a virtual reality pizza party with your friends on the other side of the world. That's less wasteful of resources than flying, you know! And how will the money work? Maybe with a blockchain.
Let's try to stretch the imagination even more! Imagine you are involved with an open source project. You want to be able to receive some donations, collectively. So you briefly consider setting up a PayPal account, then giggle. "Ha-ha! Why would I use PayPal when these smart contract blockchains technologies have been working well for many years?"
So instead you go to one of the future's several competing smart contract platforms, and you click "Set up budgeted donation contract." It asks you to input the parameters for your donation contract. Maybe you end up configuring something like "Every week, this account's balance will be allocated as follows: 25% to GitHub; 50% split among these N core developers; 25% to the fund; and 25% to the EFF."
Your fans and users can then verify the behavior of this donation contract and easily contribute to your project! Wow! So much cooler than PayPal!
How is it better than existing methods for doing whatever you envision doing with it?
What if anything, does it enable that would otherwise be impossible?
So you've got nothing?
"Ha-ha! Why would I use PayPal when these smart contract blockchains technologies have been working well for many years?"
So instead you go to one of the future's several competing smart contract platforms
Simply positing a future scenario is not an argument for it's likelihood.
Why do I need a massively parallel system with enough energy to power a city to do this when I could just do it manually, or write a script to do it and run it locally? If one of the donators doesn't trust me or my code, then they can just donate to wherever directly.
I don't even get your scenario, what value is there in me/my contract being an intermediary between my fans and these other organizations?
(and isn't GitHub a for profit company? Why would you donate money to them?)
Since you immediately reject my examples so completely, I think I won't persuade you. Good for you -- I won't fool you into wasting your time on this obviously pointless blockchain thing.
1. Payments and value management on the internet currently suck horribly compared to what you could imagine. (Insert your own imaginations here, since you don't like mine.)
2. Some attempts are being done to make payments and value management on the internet more convenient, but they are controlled by central authorities -- Facebook, Apple, Google, Twitter, Amazon, whatever -- so they will exclude, censor, limit, and control. (Just like Twitter controls your API access, Facebook forces you to use your real name, Amazon surveils your preferences, etc etc etc.)
3. Blockchains are the only way I know of to do this kind of thing without that kind of central authority, in a thoroughly open source way.
4. I can quite easily imagine a future when blockchains make it very convenient and nice for individuals and organizations to do payments and value management on the internet, in various creative and interesting ways.
Bonus point:
5. Resource use is indeed something to consider. (That also goes for, like, why are we using so many torrent boxes to seed Game of Thrones? Kinda useless! Stop wasting teh energies!) It is indeed also something that is being very actively considered by leading blockchain developers; for example, it's the major project that Vitalik Buterin is spending his waking hours on, in the form of scaling through proof-of-stake. There are highly functional versions of proof-of-stake already working for years, notably BitShares and Steemit, but they use a form of delegated voting that the Ethereum community mostly doesn't want.
" Blockchains are the only way I know of to do this kind of thing without that kind of central authority, in a thoroughly open source way."
Alternatively, the banks or payment companies are a series of non-profit foundations with protections or basic rights of customers built into their charter using regular, efficient databases with zero or low cost transfer between them. You pay a small, monthly fee to participate with your specific bank which may be waived if you do a lot of business with them. You get regular statements authenticated by the bank's HSM w/ hashes of those posted in one or more trusted, timestamping services for event bank's records are lost. Any major changes to practices might receive a majority vote by customers or people representing them from a diverse background.
So, basically traditional banks with legal incentives to not do bad stuff run by people whose background is pretty trustworthy. Many examples exist of coops, credit unions, and nonprofits that take care of customers due to good incentives. The admin overhead is kept lean per charters. The tech is vastly more efficient than proof-of-work-type schemes since it's just database transactions. Better cost/efficiency than before since new banks aren't stuck on expensive mainframes. ;) What you think?
What it can enable/improve upon according to Goldman Sachs [0]:
Assorted banking use cases, trade settlements, payments to notaries, voting systems, vehicle registrations, wire fees, gun checks, academic records, cataloging ownership of works of art.
The Ethereum alliance contains many big name companies such as Mastercard as Cisco [1]. They all believe there is real use in Ethereum
Let's think about your example and what it entails.
So you're going to a pizza party with your friends, and instead of saying to your buddy Mike who you're about to see in person and share some pizza with, "Hey Mike, here's 5 bucks, thanks!"
Instead you fire up the ol' ethereum whizbang machine, and "pay" Mike. The payment then rattles around the ol' globally distributed whizbang machine, at the SPEED of Light! powered by electricity through countless microscopic electronic gating mechanisms before finally settling into the storage media of thousands of computers around the world, where it will stay, forever.
That seems needlessly wasteful of resources and expensive.
What it boils down to, is what is this good for?
While some may have been, not everyone was confounded by the introduction of the web. Many could see obvious avenues of potential and development.
Even back in the 60s or 70s people were envisioning computer network enabled home shopping. Not sure many were envisioning a global distributed turing complete whatchamacallit with ever expanding resource needs because of no garbage collection by design.
So simply, what's the point? What do you do with it besides buy pizzas with your friends? How is it better than existing methods for doing whatever you envision doing with it?
What if anything, does it enable that would otherwise be impossible?