Some clarification is called for.
In terms of the video game market in 1982 the use of the 400/800 chipset was indeed a big advance on the Atari VCS. The two system were designed by the same engineers consectutively with the difference at the time being that the VCS would be a $300 video game while the Atari 800 would be an $1100 computer. From a game development perspective that difference was massive.
The competition was hardly any better in terms of the recentness of their chipsets. Everything in the Colecovision was equally of 70's vintage. A Z80 CPU, Yamaha/GI audio chip that was in a zillion devices, and the video chip TI had been putting in their TI-994A home computer for years when the Colecovision was released. Atari at least engineered their own coprocessors. Coleco was entirely dependent on COTS parts, which is why Japanese companies were easily able to build their MSX design as a Coleco clone and run the same carts with a pinout adapter.
The Jaguar most certainly did have a 64-bit element in its architecture in the form of its main bus. It was a 64-bit wide path which was useful for moving data around quickly and such wide busses are extremely common on graphic systems today. This is not the criterion used by most to decribe the 'bitness' of a system. By this description the Intel Pentiums would be 64-bit chips since they use a 64-bit path to memory. Yes, the Atari marketing blather was overdone but this doesn't change the fact that the Jaguar could do quite a lot of things that the existing systems when it was first released couldn't touch. As I said before, the deficiency was in Atari's funding, not in the machine itself.
The Neo-Geo was never intended to be a force in the home market. I had an interesting interview with SNK's North American manager about this. SNK's single biggest expense item was mask ROMs. They used huge amounts in their arcade boards but ordering ROMs exclusively for the arcade business meant very small production runs and very high cost per set. Creating a niche market home version their arcade machine allowed them to quadruple the size of their mask ROM orders and reduce the cost per set by nearly half. The home business wasn't a big money maker but it cut the cost of the arcade board so much that those margins were a big boost for the company. It was a great idea for a while but things changed over the years and SNK failed to keep up with those changes. They missed a big chance to get serious in the home market by taking too long to go to CD-ROM and not putting enough RAM in the systems to deal with the transition from direct memory mapping to loaded data.
I'd comment on your last paragraph but frankly, I'm not sure what you're trying to say. At what point when the SNES and Genesis were their respective producer's lead products did Sega 'quit.' If you're referring to the point where Sega withdrew from the hardware market after Dreamcast failed to reach the mainstream consumer market, what has that to do the SNES era of the better part of many years prior? If you're trying to suggest that the Genesis was a more powerful machine I'd have to disagree. Having been involved in Apple ][GS software development I have no great love for the 65816 (although the CPU in of itself wasn't what was wrong with the ][GS it very existence made the ][GS and much horror possible) the SNES coprocessor set was substantially more sophisticated than that of the Genesis.
Nintendo didn't win on hardware alone. They also had a contractual stranglehold on third party developer publishing in the US and this gave them a huge early advantage in in exclusives like Street Fighter II until Sega finally wised up and sued. They eventually settled out of court and the field was finally even. (NEC never sued because they were also a chip supplier to Nintendo and weren't willing to risk that business. So many of the best PC Engine games were never released for US consumption on the TG-16 because those games had already appeared on the NES and had a minimum 2 year period before they could appear on a competing platform, by which time the game no longer were considered worth issuing.
In terms of the video game market in 1982 the use of the 400/800 chipset was indeed a big advance on the Atari VCS. The two system were designed by the same engineers consectutively with the difference at the time being that the VCS would be a $300 video game while the Atari 800 would be an $1100 computer. From a game development perspective that difference was massive.
The competition was hardly any better in terms of the recentness of their chipsets. Everything in the Colecovision was equally of 70's vintage. A Z80 CPU, Yamaha/GI audio chip that was in a zillion devices, and the video chip TI had been putting in their TI-994A home computer for years when the Colecovision was released. Atari at least engineered their own coprocessors. Coleco was entirely dependent on COTS parts, which is why Japanese companies were easily able to build their MSX design as a Coleco clone and run the same carts with a pinout adapter.
The Jaguar most certainly did have a 64-bit element in its architecture in the form of its main bus. It was a 64-bit wide path which was useful for moving data around quickly and such wide busses are extremely common on graphic systems today. This is not the criterion used by most to decribe the 'bitness' of a system. By this description the Intel Pentiums would be 64-bit chips since they use a 64-bit path to memory. Yes, the Atari marketing blather was overdone but this doesn't change the fact that the Jaguar could do quite a lot of things that the existing systems when it was first released couldn't touch. As I said before, the deficiency was in Atari's funding, not in the machine itself.
The Neo-Geo was never intended to be a force in the home market. I had an interesting interview with SNK's North American manager about this. SNK's single biggest expense item was mask ROMs. They used huge amounts in their arcade boards but ordering ROMs exclusively for the arcade business meant very small production runs and very high cost per set. Creating a niche market home version their arcade machine allowed them to quadruple the size of their mask ROM orders and reduce the cost per set by nearly half. The home business wasn't a big money maker but it cut the cost of the arcade board so much that those margins were a big boost for the company. It was a great idea for a while but things changed over the years and SNK failed to keep up with those changes. They missed a big chance to get serious in the home market by taking too long to go to CD-ROM and not putting enough RAM in the systems to deal with the transition from direct memory mapping to loaded data.
I'd comment on your last paragraph but frankly, I'm not sure what you're trying to say. At what point when the SNES and Genesis were their respective producer's lead products did Sega 'quit.' If you're referring to the point where Sega withdrew from the hardware market after Dreamcast failed to reach the mainstream consumer market, what has that to do the SNES era of the better part of many years prior? If you're trying to suggest that the Genesis was a more powerful machine I'd have to disagree. Having been involved in Apple ][GS software development I have no great love for the 65816 (although the CPU in of itself wasn't what was wrong with the ][GS it very existence made the ][GS and much horror possible) the SNES coprocessor set was substantially more sophisticated than that of the Genesis.
Nintendo didn't win on hardware alone. They also had a contractual stranglehold on third party developer publishing in the US and this gave them a huge early advantage in in exclusives like Street Fighter II until Sega finally wised up and sued. They eventually settled out of court and the field was finally even. (NEC never sued because they were also a chip supplier to Nintendo and weren't willing to risk that business. So many of the best PC Engine games were never released for US consumption on the TG-16 because those games had already appeared on the NES and had a minimum 2 year period before they could appear on a competing platform, by which time the game no longer were considered worth issuing.