In computing, the Computing-Tabulating-Recording company (later IBM) produced the first printing tabulator, the beginning of commercial data processing.
Those 50 years held the greatest lifestyle changes in history.
[1] https://www.youtube.com/watch?v=LetlcmqZFyA [2] https://en.wikipedia.org/wiki/Wirephoto
More generally, though, "email" was called "mail". In cities, there were often two, and sometimes three mail deliveries daily. It would be possible to carry on a round-trip correspondence over the course of a day.
The precursor of online shopping was called "catalog stores" -- Sears, Roebuck & Co. was the canonical example. I just caught an item over Ansible last week that Sears had discontinued its catalog in 1993, the year before Amazon started.
Robert Gordon, economist at Northwestern University, has just published a book on technology and growth, The Rise and Fall of American Growth. In it he notes that the typical urban home of 1920 was connected to five "networks": water, sewer, gas, electricity, and phone. I'd add postal mail and the highway, railroad, canal, and shipping networks as well, though those had a "last mile" component of individual delivery. Still, this largely offered similar capabilities to today's Internet.
Email was postal mail.
Online shopping was catalog shopping.
References, information, and entertainment (as well as advertising, without tracking or pop-ups) were performed via books, encyclopedias, newspapers, and magazines. Phonographs provided audio, as well as over-the-air radio broadcasts. Libraries provided extended access.
Latencies were higher, but the throughput of even a horse-drawn cart is pretty good.
but there was not really any concept of store and forward like you have with texts and email.
Notice how they had to tell the operator to make sure there was a clean line, then manually sync the two devices.
I guess you could claim the telegraph network had something akin to store and forward, as each operator would note the message, and then relay it down the chain.
Howard Krum finally figured out how to do sync right in 1919, when he invented serial stop-start async as we know it today. In 1924, he came out with the Teletype Model 14 tape printer, with the first reliable mechanical async decoder. That design was used into the 1960s, and most computer serial ports are compatible with it. (45.45 baud, 5 bits, no parity, 1.5 stop bits.)
Telegraphy was store and forward, but manual at first. By the late 1920s, punched tape was used, but someone had to manually look at tape, tear off each message, and put it in the proper outgoing reader. This was manual Sendmail. After WWII, automated forwarding centers were built, with paper tape punches connected to readers via a bin as the storage mechanism and telephone-like switches to change connections. These were hugely expensive to operate for the amount of data transferred, but it was essentially email, with electromechanical technology.
(I restore Teletypes as a hobby. Here's a Model 14 working.[1]. If you want to see some in action, they'll be at the Clockwork Alchemy Steampunk Convention, Memorial Day weekend, in San Jose.)
(Edit: in fairness, that would be more like "land line texting" than email.)
That said, it seems most of them were operated over dedicated networks. And the one i can find a reference to that was operated by AT&T had their own area codes etc.
All in all i get the impression that it could be done, but likely only so at great expense. Consider that having a ticker tape printer in your home suggested you were someone with great wealth that was watching the stock market 24/7.
Even having a telephone was not chump change. My parents talk about having one phone for several households, with the bell patterns used to indicate who it was for.
I suspect we should not underestimate how much the IC and packet switching has driven down cost of installation and operation.
But suppose someone like Edison had gotten hold of the 1920s transistor lab experiment and tried to make it better by trying many different materials. (That's how he got a usable, cheap light bulb filament; thousands of materials were tried, ending with carbonized paper.) They would have tried selenium (selenium rectifiers were known) and galena crystals (crystal radios were known). They probably would have gotten a point-contact transistor out of galena, which would have led to germanium. It would have been unreliable and heat-sensitive until germanium purification was figured out. They might have tried grey tin, which is a semiconductor but a dead end. They probably wouldn't have tried silicon, because you have to grow silicon crystals first, which is hard and wasn't done until 1950.
So a transistor radio might have been demoed by 1930 or so.
But yes, the 50 years from 1870 - 1920 essentially created the modern world as we know it. Shy television, many plastics, antibiotics (though massive healthcare improvements had already been made), nuclear fission, lasers, transistors & computers.
Robert J. Gordon's The Rise and Fall of American Growth does a stellar job of detailing the various impacts of technology on the period 1870 - 2015, divided into roughly three fifty-year periods: 1870-1920, 1920-1970, and 1970-present.
The combination of electric motors and internal combustion engines seems most specifically powerful.
It seems, however, that this theory is not very well supported except for the general impression one might get reading about his behaviour in other situations - Newton doesn't seem to be the humble type.
There was also the backstory for Alpha Centauri. Firaxis' sci-fi Civilization where the landing of the space ship goes anything but smooth.
I'd also strongly recommend the subsequent programme, The Day the Universe Changed, which follows a similar, though temporally arbitrary, connect-the-dots motif.
The two sequels to Connections aren't as good, though C3 is the better of the two.
Also: if you're interested in the background of technology, look up the story of Joseph Needham and his extended exploration, Science and Civilisation in China. Begun in 1954, it remains in process, with 27 volumes. Needham himself died in 1995. Simon Winchester's written an excellent (and much shorter) biography of Needham.
So, if your starting with the correct raw materials and enough food you could probably get to early CPU's in 10 years or so.
I think that example demonstrates the opposite. Getting raw materials is not a problem; it's so easy that people thousands of years ago were shipping raw materials thousands of miles when there was a need for it.
It's knowing the need and having the need which is the hard part... Tons of silicon can be found on your local beach and have been forever. But we only got computers a century ago.
Sure, now we can extract gold from sea water and it's just a question of cost. But, in a collapsed world your much more limited.