Then why switch now? ARM is not natively compatible with today's software.
I will give you some history, to make this more clear. When Apple and (for similar reasons NeXT Computer), adopted the 68000 (family), it was a cleaner architecture than Intel’s competitor (the 8086/80286). The previous generation (the 6800) had a similar market share to Intel’s previous generation (the 8080), so the choice was reasonable. Many workstation companies adopted that family, but IBM and its clone makers having chosen Intel, gave them a commanding lead in volume.
When Apple wanted to switch to a faster architecture, it made sense for them to work with partners (the AIM consortium) to build the PowerPC. Unfortunately, Motorola (and later Freescale) had much more success with the embedded systems market and Apple was not big enough to support development of a super fast chip on its own. IBM did use the PowerPC in its own systems, but only in mainframes, so performance per watt was not important for them, again a problem for Apple.
Various other vendors did get Microsoft to port Windows to their architecture (MIPS, PowerPC, and DEC Alpha, among others), but while they were all faster than Intel, none ever had enough market share to get people to port to them, many had endian-issues (making porting harder), and the most complex applications have lots of platform specific code, ensuring they would never run on anything other than Intel hardware.
At that point it made sense for Apple to move to Intel.
Steve Jobs decided that Apple should never be in that position again (not controlling its own destiny) and began building a world class silicon team. Apple’s volume in the iPhone, iPad, Apple Watch, Apple TV, HomePod,
etc. is large enough support this team.
The reason this makes sense for Apple and does not make sense for anyone else, is that Apple controls all parts of its own platform, and can make both make it easy for people to port, while making it impossible for them not to port if they want to remain on the platform (“burn the boats”). For anyone on the platform, who is on Apple’s tool chain, porting is trivial. Where there may be a small set of developers who leave, that really means that they had left before and did not bother telling anyone (in other words, they were not doing any work to support the product, just collecting checks).
No one else has had the same incentive structure to do this for desktop/laptop systems. Nor has anyone else had control over all parts of their stack.
An ARM chip marker, would have to build a chip, get a system maker to build a system around it, get them to write drivers and work with Microsoft to ensure there is a port. Unless they achieve great volume, no one will port to their architecture, and if no one ports to their platform, they invested a lot of money for no reason. Why target that market which is shrinking in importance
vs. the mobile or embedded markets?
As for ”native compatibility” your statement is clearly wrong. Apple’s biggest markets all use ARM, so the number of systems that use ARM
vs. the number that run iOS/iPadOS/tvOS/watchOS.
At this point, people can rant, leave the platform or wait and see. Unless one has an urgent reason, waiting and l cannot see any good reason for anyone who does not need to buy a machine today to buy one now or switch, when they can wait to the fall and see what happens.
developer.arm.com
Apple has been remarkably consistent in their margins, that means that we are likely to see either much faster machines at the same price points, or machines with similar specs at lower prices (or some combination somewhat better specs, somewhat lower prices).