The Amiga was in a downward spiral after the A3000. That was in my view the final true Amiga that showed progressive hardware development and forward thinking ideas because the development of that system had the money spent on its development to allow it.
After that Commodore was starting to play catchup with the 16-bit consoles in gaming terms, and with the PC in computing terms. The A4000 was a cut down shadow of its original concept. Still a very nice and desirable Amiga at its launch, but not the next step in Amiga development by a long way. And the A1200, while a great Amiga and the most used these days, was also a cut down budget restricted design compared to the original ideas. And the CD32 was just their last ditch attempt to claw back into the gaming market by using their A1200 CD drive development.
However they were not alone. Atari were in a similar situation with the Falcon. The final released system being a cut down version of its full original design, and failing far quicker than the Amiga and Commodore.
I think the biggest problem was the speed at which computer technology was progressing. Not the dodgy dealings of the companies involved. At the start of the 80's 8-bit systems such as the C64 and Spectrum had very long shelf lives. Although some different models were released for certain makes containing more ram or a different keyboard, the core remained the same and the companies got used to having a static stable platform that was easy to mass produce for years with little change to the design, but which sold well. That all changed in the 90's. New hardware and development of existing hardware was getting faster and faster and it just wasn't possible to use the same long development cycles they had got used to.
This is where the modular mass produced nature of the PC market suddenly came into its own. The ability to chop and change the spec of a computer to suit each user, rather than buying a fixed platform and then trying to bolt extras on.
For a fixed platform gaming now had consoles. These retained their long shelf life and minimum continued development cost. Commodore saw this and tried to jump ship and become a player in the console market with the CD32 because that market fitted the development model they had become used to with the C64 and then the A500. What they failed to understand was the shear size of the budgets the console developers had to play with. Something Commodore didn't have.
The biggest failing of Commodore was as is so often mentioned that they just didn't know what to do with the Amiga. How to market it. Who to market it to, and what markets it fitted into.
Apple only survived because they did know their market and stuck to it. They never tried to seriously enter the gaming marketplace. They knew their core market was in DTP and design and so kept on making robust workstation like systems for the professional marketplace. A decision that is what allowed them to survive. The Mac was however a different market to the Amiga as it was restricted more so to certain professional userbases, so its marketing model cannot directly be compared.