>> Plenty of failed hardware in Amiga history. Including some which some
>> of the highest regarded Amiga engineers worked on . So It can even
>> happen to the best . AmiJoe anyone ?

> they failed for exactly the same reasons. PPC is dead.

That's not the reason most of these projects failed a decade to two ago. As for the AmiJoe in particular, you can read up on its failure there.

> For PPC accelerators, nobody had the technical knowledge of Phase5

Thomas Rudloff, the AmiJoe developer, came from Phase 5 where he co-developed the CSPPC and BPPC, for instance.

> Phase5 ran a business making accelerators for Macs. Nobody else did.

This claim is laughable. Phase 5 were one of at least about 15 companies doing so, and by far not the first one at that.

http://www.macinfo.de/tuning/cpuupgr.html
https://everymac.com/upgrade_cards/by_manufacturer/

> For full boards, it's because of lack of reliable support chipsets.
> Apple designed its own and didn't sell them on. It didn't help that
> Freescale went the SoC route.

The SoC route at least potentially helped with the alleged lack of reliable support chipsets ;-) And I think there were reliable G3/G4 support chipsets made by IBM, Motorola/Tundra and Galileo/Marvell, as well as G5 support chipsets made by IBM (as used by Apple) and Marvell. After all, how was it possible to build high-reliability systems with these CPUs if that wasn't the case? It's just that there were never reliable support chipsets featuring AGP (which was deemed a requirement for desktop systems pre-2003), as they went straight from PCI(-X) to PCIe in 2005/2006.

> A-Eon got in on the action.

Don't forget ACube :-)

> "Hey, this board isn't supported well enough. Time to make a new one
> and support it even less!"

You would be surprised by Costel "Cyborg" Mincea's os4welt.de comments on A-Eon's board release strategy ;-)