ARM processors market is too much fragmented, every manufacturer implements what he thinks it's the best and things will not be standard soon.
Usage of ARM processors covers a extremely broad market from SSD controllers and embedded contexts up to server processors. Broad coverage is very different from fragmentation.
If narrow to the phone space it isn't all that fragmented. Qualcomm , MediaTek , Samsung, Apple, and HiSense (and whoever steps in to replace them as they are elbowed out ) pretty much cover the whole market. The attached modems ( or not) vary but the instruction set is pretty much standard AArch64. There is a standard. It is the one Arm laid down.
There is some mild variances over extensions and who is most up to date with the standards definition.
For the classic Personal Computer it probably won't be all that fragmented
Read this introduction to the Arm Cortex-X Custom Program, outlining what the new program entails and providing details about the new Arm Cortex-X1 CPU which is part of the program.
community.arm.com
If ARM can find a substantive enough group to keep funding Cortex then by X2 , X3 , X4 there should be a very substantive standard base for a laptop processor for the WinPC market. If Microsoft jumps into the group all the more so.
[ There is some coupling to the Neoverse cores here also. Which again if Amazon, Marvell, Ampere keep licensing it. That too will be more collective R&D investment. ]
I don't even think that this architecture will land on consumer PC market that much (at least for the upcoming decade...)
That is at least more up to Microsoft ( work on x86-64 conversion and a stable OS , forward looking API to port to ) than it is to ARM or the implementers.
But if Microsoft and implementers screw it up with sloppy execution you are vastly deluding yourself if think Apple is going to jump in and save them from their incompetency . Apple milked Windows bungling Vista and Win8 for quite a long time. Screw it up on hardware? They'll milk that too.
The bulk of the Windows market is relatively safe because Apple doesn't want to sell "everything to everybody". Didn't before. Do not know. Extremely likely wouldn't try to follow that path in the future. Apple will probably use Apple Silicon to sell into a subset of the more profitable subsegments of the Personal Computer part and ignore the rest.
EDIT: things will be easier for Apple as they don't care about legacy stuffs and they move forward: don't care about legacy users that will sooner or later upgrade to newer tech, like what happened with Metal and 64-Bit.
Again Apple doesn't want to sell everything to everybody. The low-mid range "box with slots"? Nope. Luggable weight mobile desktop replacement? Nope. Mega workstation with two CPU packages? Nope. Entry Chromebook competitor? Nope.
Which in part leaves them with far , far fewer processor packages to implement. Fewer, more focused implementations.
Finally, highly questionable whether Apple Silicon can even boot another non-Apple OS with the standard firmware. "Don't care about legacy stuff" isn't just up at the OS outer layers. They never were big fans of BIOS . And when UEFI turned out in large part a way to pragmatically keep BOIS alive much longer; probably aren't big fans of it either.
It isn't like Apple doesn't give decently long transition periods for folks. But Apple isn't waiting around "forever".