I also am interested into how this shaped the set up of the typical recording session. How did this affect the way in which sound engineers would rig up the equipment and if so, what were the knock on effects of this?
Thanks :)
During World War II, some technically-minded people listening to German radio were puzzled: was Hitler really demanding that the Berlin Symphony Orchestra play Beethoven symphonies at 3am in the morning? It was a puzzle, because the sound was pristine, without the clicks and pops you get from a vinyl disc - at that time, the only option for playing pre-recorded material that the Americans and English knew about. They couldn't figure out how the Germans were doing it until, after the war had been essentially won, and the Allies gained access to the premises of Radio Frankfurt in Bad Neuheim, which had been broadcasting the Beethoven symphonies. One American Major who was a classical music fan and curious about German radio, Jack Mullin, decided to head to Bad Neuheim, and discovered that they had a new method of recording to Magnetophon - to magnetic tape - which crucially had an AC bias that enabled almost pristine recording quality.
Previous to this, recording had essentially been straight to vinyl disc, which was more limited in a variety of ways - once the groove had been laid down by the transcription needle, it was that way forever. However, magnetic tape could be altered - you could tape over it.
After the war, Mullin worked with the company Ampex to replicate the technology they'd seen in Frankfurt. Initially using reels of tape taken from Bad Neuheim, and on prototype technology, they pre-recorded radio shows for Bing Crosby, a major star at the time; after Crosby gave the company $50,000 with no strings attached to perfect the technology, they released the first commercially available American tape recording devices in 1948, which basically instantly became the standard.
Crosby gave an Ampex tape machine to his friend Les Paul, a jazz guitarist (and inveterate tinkerer who worked with the Gibson guitar company on the guitar brand that bears his name). Paul, like Mullin, had noticed the German broadcasts at 3am; he was working at Armed Forces Radio in Europe during the war, and he couldn't figure out how the Germans were doing it. So upon getting an Ampex tape machine, Paul was very keen to play with the technology, and discovered that 'overdubbing' was possible; you could record yourself playing along with a previous recording - as a singer you could harmonise with yourself. So you get recordings like Les Paul's recordings with Mary Ford (e.g., 'How High The Moon' from 1951), which were exceptionally popular, and which showed Paul playing several guitar parts at once, and Ford harmonising with herself.
This was effectively multi-track recording in one sense - there are multiple tracks of music recorded one after another - but it's not multi-track recording in the modern-sense, because there weren't multiple separate tracks of tape for each instrument - there was just one tape, with different performances literally dubbed over pre-existing performances.
An Ampex employee, Ross Snyder, heard these Les Paul and Mary Ford records and thought that the overdubbing method they used lead to a decline in sound quality, and so he aimed at developing a tape machine that had multiple tape heads that were in sync with each other; this was a complicated project with a lot of technological constraints - getting different tracks to line up with each other and making sure that the tape was at the right speed etc was something of a problem. But in 1956, the Ampex Sel-Sync was put on the market, an eight-track recording device. The Sel-Sync itself was generally seen as impractical for recording studios (it was enormous and heavy and glitchy), but other multi-track recording devices came into vogue in the late 1950s.
In England, someone at EMI who had seen the Magnetophons at Bad Neuheim had also had the same idea as Mullin, and they had unveiled the EMI BTR1 in 1947, a single track tape machine. After some improvements to the design (the BTR2 in 1952), and due to the coming demand for stereo recording, EMI put together a BTR3 in 1956 which had two tracks; however, these were not used for multi-track recording, per se, until just before the arrival of the Beatles in 1962, when a modified BTR3 at Abbey Road dubbed the 'Twin-Track' was used in order to do rudimentary multi-tracking. However, in late 1963 the Beatles began using a Telefunken M1 four-track recording machine made in Germany (which had been on the market since 1957, but which EMI experimented with extensively before allowing its use (Telefunken had also built the Magnetophons in WWII). In 1965 the Beatles started using a Studer J37 four-tape track recorder which was more suited to studio multi-tracking experimentation, and (a modified version of which) was used on Revolver and Sgt Peppers - the albums where multi-track recording most obviously became an artform in itself, with musicians creating sounds using the studio and the multi-track recorder as an instrument in itself.
Edit: To deal with your second question - for a sound engineer it would have depended on what kind of music they were recording. Firstly, in the mid-to-late 1950s, Frank Sinatra was still recording basically on a single microphone, with the orchestra simply softer than him in the background. In contrast, in more or less the same time period, there was a maze of cables on the floor of the Motown studio at Hitsville USA (thus why the studio was dubbed the 'snakepit') indicating that there were leads and microphones going to different instruments set at different volumes - recording desks that mixed the volume and frequency of different instruments with each other already existed and were more common in pop rather than in orchestral contexts (especially because the levels of rock instruments could vary so much, and needed further control). Basically, once multi-track recording became a thing, it enabled engineers to record as pristine and perfect a sound of an individual instrument as they could, and multi-track recording multiplied the time it took for a band to record a song in the studio, with each part being looked over in detail to try and perfect it; often, when a band was playing together before the era of multi-tracking (or in the early years of multi-tracking), the band recording was basically it, perhaps with overdubs. But as things progressed, musicians would first lay down a basic track and then elaborate on it later, or record instruments one by one. So this meant that the original basic track didn't have to be as perfectly balanced and mixed as it had to be in the pre-multi-track days, as it was likely going to be elaborated on, and that musicians and producers could think more carefully about how the music sounded on record in context, and arrange the music accordingly. Geoff Emerick, who was the engineer on much of the Beatles' records, speaks of his disillusionment on working on the Beatles' stuff in the later era when multi-tracking was a thing in his book Here, There And Everywhere - he thought that the broad array of possible options and the pursuit of perfection led to a sort of paralysis, and made working in the studio with the Beatles much more difficult.
Sources: