In my humble opinion there was no "worst move". It was a series of related moves that caused CBM to fall...
As many companies before and after them they had no real strategy for their products. I think this is mainly to do with bad leadership. Too many people have their say in the organization and you end up with competeing productlines and no real market-strategy.
Just look at their products: The Commodore 64 line and it's derivatives was kept going way too long. Just because the silly little machine made it to the Guiness Book Of Records they didn't have to keep it going for years after the world had switched to 32bits on most every computer-platform.
Then they had a PC (x86) based lineup. These computers weren't all bad compared to the competitors at the time, but as I recall they were not competitively priced.
The Amiga was never really branded - it was a games machine in Europe, a video workstation in the US. And Lord knows what else... It was never communicated to the world what the machine really was. Now we might see this as great because we know what the Amiga really IS, but to the average consumer a computer with no message or identity is just that....
And to top it all off, CBM had a server-lineup. Who knows what they intended with that one. They didn't want the Amiga to be a UNIX machine because it would confuse the buyers. They didn't want to "OEM" it to SUN for whatever reason... But they poured R&D-funding into the CBM 900 - again with no clue as to where they wanted to go with it. And that machine WAS a UNIX-based contraption.
Do you see what I'm getting at...?! :-?
Commodore was all over the place - and I have seen up close and personal what that does to a company!!
The Amiga, of course, suffered from all of this.
There was noone at CBM that was tough enough to kill the 8bit product-line and sell off the PC-lineup.
So the Amiga had to compete with it's mothercompany's own products - and trust me when I say, that is A LOT harder than having to compete with other products in the real world...
CBM should have focused solely on the Amiga.
They were just too small to compete with IBM and the other PC-monsters.
With focus on the Amiga-line they should have developed the machines identity within the market it targeted. Not try and have it compete with Nintendos, PCs, Macs and whatever.
Home computers, designer workstations, Video workstations - and maybe a relabeled UNIX-machine with a server line to hook it up to... Come to think of it, this is where the CBM 900 would fit neatly into the equation.
It would have been a great world today if companies ran Amigas at the business end, and CBM 900 derivatives in the server room. Mmmmmmmmmm....
If I only had the chance to run Commodore for a few years... But sadly I was only 12 years old in 1988, and I had no idea about how to run a business... :-D