Technically the Atari 8-bits had a 128 color palette, not 256, but nonetheless they did have pretty much the best color palette on the home computer market for several years. While I respect that the C64 edges the Atari 800 in several technical respects (the sprite hardware is more flexible, the sound chip is superior... although the Atari had a faster CPU and the ANTIC/GTIA combination for video had some interesting tricks up its sleeve so it's by no means a walkover) I've always thought that the Atari 800 ports of games tend to be "prettier" than the C64 version. But, yeah, the first Apple computer to be as "colorful" as the 1979 Atari 800 was 1986's Apple IIgs. (Technically Commodore's "264"/Plus/4 had a similar palette to the Atari 800, but it was an utter failure when it came to market in 1984.)
MIDI. ST's were really, really good for that. Some anal-retentive synth freaks use them to this day.
The other thing the Atari ST was good for was pointing out just how comically overpriced Apple's B&W Macs were. ($799 list price for a 512k Mono system with a single floppy vs. *drum roll*... $2795 for a Fat Mac.
One could sort of make the case that Jack Tramiel ultimately managed to torpedo *both* Commodore and Atari in the end by making each company a victim of its own success at cutting prices (and profits) down to the naked bone. If you concentrate too hard on marketing your products as "cheap alternatives" to an established premium competitor you run this risk of simply having your products perceived as cheap knockoffs regardless of whether they may actually be technically superior. (In at least some regards, anyway.) The other problem, of course, is that if you don't make enough profit to plow some money back into R&D you're eventually going to hit a wall; it's interesting that both Atari and Commodore basically managed about 2 1/2 generations of their platforms before they were crushed. IE, they both started with the initial, exciting 16 bit model, followed by a "workstation-class" 68030 machine, and then... a few years of stagnation followed by a last gasp run of machines that were cheaper rehashes/remixes of the previous two generations. (Yes, we could have a huge argument about whether the Amiga 4000 was a cheap rehash of the 3000 or a big leap forward at this point, but it's inarguable that the standard configuration actually included an *inferior* CPU to the 3000. Same goes for the Atari Falcon vs. the TT.) Motorola giving up on the 680x0 series of CPUs sure didn't help; Apple had the resources to follow in the footsteps of UNIX vendors and jump to a RISC platform, Atari and Commodore were just screwed. It's at *that* point that Commodore lost the plot. And let's be clear on this: it's not because they had bad engineers, it's all about bad management.