Bad marketing - it's not understood
Thunderbolt is PCIe over a cable. Your internal expansion cards are now external. Enclosures exist that let you plug in whatever cards you like, even if you have no expansion inside your actual computer. So your laptop dock just allowed you to have a full spec desktop while docked, complete with powerful (but external, now) graphics card, arbitrary drives, monitors, other expansion ports and so-on.
USB 3 is a completely different technology, a distinct bus/interface, compatible only with itself, with huge piles of driver software layers implementing both the USB 3 and older parts of a standard that was a bit of a creaking mess even back when USB 2 was added. All these years later, there's a good reason vendors are moving to things like SPI to connect internal low bandwidth interfaces rather than dangling them off USB - it's just too slow and software-heavy, even for USB 1. Things got a lot worse when USB replaced PS/2 for those devices, and now, finally, SPI sorts that out.
So what of Thunderbolt? Why is it so rare? The one obvious area where Thunderbolt fails technically is the hugely expensive active cables. Even if they were cheap, active cables are an obvious point for things to fail (and fail badly). But even putting that aside, Intel and Apple decided to push the technology via transfer speeds - a peeing contest. Rather than just say "It's way faster than USB 3" and *then* focus on all the unique and really interesting aspects, they just said "It's way faster than USB 3" and pretty much stopped there. Customers are left in the dark. From their perspective, it's just a funny looking port that has really expensive peripherals. The genuinely remarkable opportunities it offers for new form factors - a much better approach for a hybrid desktop->laptop->tablet kind of affair for example - have been overlooked.
Net result: Very little uptake, no economies of scale, specialist peripheral vendors only, extortionate prices.
The forthcoming updated Mac Pro shows what can be done, but I fear too little, too late and too specialised. I'd love to upgrade my now very old Mac Pro with the new machine, but even if I can afford the base unit, I almost certainly cannot afford the peripherals I'll need for the expansion cards and drives I'd need to carry over.