An aside, but the Mac ROMs were never "cloned" in the short-lived PowerPC 'clone' era. Rather, Apple opened their specifications, and basically gave everyone a license to a reference ROM. Thus, when Apple chose to reneg on their contracts and rescind such licensing, the clonemakers were forced to close up shop. (Presumably companies like Pios could've taken Apple to court over breach of contract- I assume someone must've- but as we can see, it would've been hard to get a ruling that would keep Apple from 'breaking' future OSes from running on the hardware, and apparently no one did.)
As all the alternative manufacturers were 'tainted' from their contact with the Apple intellectual property, none of them could step forward with a "clean-room" implementation- something Compaq took great pains to achieve in producing the first IBM clone.
I recall one 68k Mac emulator (for OS/2, no less) that stepped around the ROMs entirely by simply emulating/translating the entire MacOS API (not unlike the "Project Odin" for Win32 compatibility, but with an added element of 680x0 emulation). There was also a short-lived 680x0 Mac laptop made by "Kangaroo Computers;" I assume they were also licensing their ROMs from Apple.
Now, this all relates to the Amiga vs. MorphOS case in that there is a (potentially unanswerable) question as to the 'cleanliness' of some of MorphOS's implementation. If they got their code through AROS, they're clean; if they saw some AmigaOS source through H&P, they aren't. Whether recycling of binary elements from AmigaOS during testing (as the icons demonstrate, if not the Workbench) constitutes fair use is another interesting debate.
