If you're asserting that SCSI and SATA should unify, I think you would be better suited to unifying other standards that make no sense why they're fragmented. For instance, there's a standard for rendering web pages, but there's like a half dozen ways to do it. This is a nightmare for web developers because they have to make sure each of these rendering engines and bloat their code at worst case 5-6 fold. At least SCSI and SATA are two incompatible standards and both of them have been around for a very long time (SATA is just an extension of ATA after all). And they're hard standards rather than soft ones web protocols specify, so pretty much every SATA controller is going to look the same.
Also, this is a pretty tall claim:
I know a lot of people weren't affected by these problems, but a lot were as well. SATA is the CAUSE of these problems, and the bait tempting hard-drive manufacturers to lower the bar again and again.
Do you care to provide some hard evidence rather than talking out of your rear? And how do you know that through a combination of commands that I can't make the SAS interface poop out? POSIX operating systems are more robust than Windows NT based ones by default, but you can easily break a POSIX OS with the right commands, just as you can a Windows NT OS. Can we blame OCZ's SSD issues are because of SATA?... No, it's because OCZ wrote crappy firmware that controlled the bare-metal communication to the storage medium
In any case, the end point is that a lot of these errors the enterprise and scientific computing are worried about happens so infrequently to the average customer that it's not even worthwhile. For that extra 1 part per million chance out of a 1 part per billion system. And as pointed out, do you want a system with all the trimmings? Get a workstation board. Most of us are content with what we have. And if I'm not going to see an appreciable difference for "insurance", then I'm not going to be happy paying out more money for a part that will eventually "lower in cost".