The worst possible move was having idiots run C= that didn't have a freaking clue or plan. All they did was react to events.
Like any aircraft accident, there wasn't a single blunder, but a series of small blunders that acted like a domino effect that leads to a catastrophic accident. C= should have had a premium customer service beyond a decent base customer service. C= should have trusted their engineers and accept them as a full fledge member of management team. C= failed to accept they needed a baseline nitch market (AV market), worked with AV companies in a unified developement push.
C= failed to keep the Amiga current with x86 (486), they should have spent the money on porting AOS to x86 and setup and spun off a x86-AOS company with minimum cash drain plus turn a blind eye to piracy. This is exactly what M$ for decades, whine about it and let piracy push M$ into ever office and home. Let the tiny AOS/x86 company fail after maximum saturation and allow the clone OEM to license at a really cheap rate AOS. Most of Amiga folks would never go to x86 back then so there is little worry about losing us to x86 sales.
C= also failed to buy at discount rate, failed AOS software houses' IP. Had C= started to spend minimum cash for the better titles, they could once again accessed cheap coding markets that were beginning in India. Port the games and decent apps to x86 and offer it to the x86 OEMs for minimum money in addition to AOS allowing the OEM to resell it on CD as value added content. Doing all the above for say a fifth of what M$ wanted would given C= a foothold in the markets across the board.
Dammy