All the last of the Amiga's, etc had USB and many of the other industry standards, it wasn't openenss that allowed IBM (PC) to win.
Part of it was .gov contracts which required second supplier so the Intel/AMD/etc partnership was formed to have a second supplier for the main chips.
The architecture was open. The only (read literally, ONLY) closed part of the IBM PC, the
only thing IBM actually
created, was the BIOS. It was all open reference stuff. It's like they took a "how to build a generic computer for dummies" kit, put it together and slapped the logo on it.
There was nothing open about it. Compaq enabled that by pouring a metric crapton of money into a clean room clone of the BIOS. *That* is what enabled the flood of clones. A whole heap of people looked at it and went "hey, we can do that". From then on there was no need for a clean-room implementation as they could pick and choose bits from both and it became very quickly impossible to figure out what bits of the mutant came from which source, and therefore pretty much impossible to prosecute.
Yes, the compatibles did start with Compaq, but so did the actual triumph of the PC. Until Compaq it was a reasonably successful business computer, but nowhere even remotely close to market dominance. IBM never truly profited from the rise of the PC. The PC rose by virtue of having been "stolen from" IBM and put in "public domain".
And as I said, it WAS open 99%. The BIOS wasn't open, and how hard was it to create a cleanroom reimplementation of that - as evidenced by all the companies who did this, it wasn't too hard. Note, this could never ever have happened to the Amiga or even the C64. Nobody could have made a 'C64 compatible'. A Macintosh compatible, that was beyond dreams. Copying an entire architecture without infringing on patents, and cleanrooming a few kilobytes of system code are lightyears apart.
Software was the real big win, Microsoft software. The first question over any Amiga/Atari, etc sale was can it run MSOffice/excel/word.
You're confusing time periods and causality. Early Microsoft software was demonstrably inferior. Excel was inferior to Lotus, Word was inferior to WordStar.
It was the other way around. Microsoft software only became popular due to the rise of the PC, which was due to the availability of cheap compatibles. The reason that people asked
in the dying days of Atari or Amiga if it runs MS Word was that PC compatibles have already won the war, and everyone ran MS Word at work by that time.
The thing Microsoft did contribute to the compatibles ecosystem was being an independent OS vendor. If PC DOS was actually an IBM exclusive product, the lack of a compatible OS could have prevented the rise of 100% PC compatibles. Since PC DOS was only a rebranded MS DOS... well we know the story.
Developing for multiple platforms isn't that big of a deal, look at firefox, chrome, thunderbird, ApacheOpen/LibreOffice, KiCad, etc. Running parallel versions for win/mac/linux isn't much work that running parallel for win/mac/linux/amiga.
It was back then. Besides, don't compare running different OSes on the same architecture and same CPU to having different architectures, different CPUs, etc. Yes, the GNU ecosystem can live on a wide range of architectures thanks to the immense work put into their optimizing C compiler(s). But that took immense effort.