Re: what a lot of people..
The Amiga and ST blew PC hardware into the weeds
The Amiga didn't have an MMU as standard, hence memory protection between workbench applications was crap.
The Atari ST didn't have a blitter.
All of the rose tinted glasses in the world will not change the fact that there was NO decent commodity hardware available at a decent price back then.
I actually lived through this era and used these machines - all of them, IBM-PCs, Macs, Amigas and Atari STs. All of them had nice features. None of them had everything.
And because it hasn't been mentioned so far, a vote here for Acorn's Archimedes (1987 IIRC), yes it was a little later than the Amiga and the ST but it falls into a very similar category. I lived through the era too and did my fair share of typing listings out of magazines and writing my own software because I'd used most of my savings on buying hardware.
My view of the situation has long been that the reason MS "won" was nothing to do with the company itself nor its products, but was everything to do with the hardware. Not one of the other, potentially competing, systems turned out to be as "open". As soon as Compaq had reverse-engineered IBM's BIOS it became impossible to bolt the stable door and get the cat back into the bottle. The basic hardware itself was hardly more than Intel's application circuit for the 8086 and suddenly it became relatively cheap and easy to get "good enough" computer hardware from a number of suppliers.
As for an OS, only MS had something ready-to-go on that hardware. They were "in the right place at the right time". Can you imagine Commodore getting the Amiga OS working on that stuff? RISC OS? The key point is probably that they didn't want to. Their business model revolved around selling a combination of hardware and software and differentiating their products in ways that your average business user didn't understand and frankly didn't care about.
Yes, there were "killer applications" on each system; the ST's built-in MIDI, the Amiga's video circuitry (the Toaster) and, of course, Sibelius under RISC OS, but there was nothing intrinsic about those systems that meant that only they were suitable for those applications. Eventually the "PC" caught up.
And once home users began to understand what computers would be most useful for, they ended up buying Canon Starwriters or Amstrad PCW machines - in many ways "appliances", in a way more closely-related to tablets and smartphones than to modern PC-type computers.
Apple's flirtation with clones sort of missed the point, but I was well out of the Apple ecosystem by then so I can't really comment.
In a way I miss those times, but with a fleet of several dozen Raspberry Pis, a fistful of Arduinos and a copy of BBC BASIC for Windows (thanks to the marvellous Richard Russell) it hasn't completely disappeared.