Apple didn't write an OS from scratch either. iOS was really a fork of OS X, and followed on the experience Apple & NeXT had in porting NeXTstep around various architectures. Something that MS also had a lot of experience with, with NT.

Apple won/created this market on fit and finish, excellent marketing, singular focused execution, and supply chain genius.

With wince, Microsoft was running around farming out their OS to multiple 3rd party vendors, trying to create a commodity marketplace for their OS on other people's devices.

Apple won on vertical integration, created a focus product, and profited off the experiences and reputation they had with iPod in order to make a winner.

But people also forget that at this point in Apple's history, it was kind of... make or break. The iPod and then iPhone were huge gambles, that paid off in spades. If they had failed Apple would have continued to be a minority PC producer with poor profit margins and a reputation for being "weird" and "niche." Even after the success of the iMac in the late 90s, they were still in quite a bit of $$ risk.

The intense vertical integration and the product focus in the iPhone could also have failed. It was risky but it paid off for them. MS (and others) had a lot more to lose, and was more conservative.

Apple had giant profit margins on Macs even back in the iPhone days.

If that was the case android would also be in the sinking same boat. Vertical integration isn’t what made iPhone and android.

I like this hypothesis better than mine.

I believe WinCE was a "new" (or at least different) OS. it sort of had Win32 like APIs on top of it, but the core was fundamentally different than NT. They only sort of got things moving in the right direction (only to then abandon it) when they had their mobile platform on top of the "normal" (pared down) NT stack.

* * *

They likely had to promise that it would run Office as well.