Our choice of SSDs was limited to what was shipped in each sealed device from the factory. So if the SSD they chose was not as fast as the competitor's, that's the choice they made and they have to live with it.
It also limits the choices of the consumer. You get what you get unless you want to crack the case open and void the warranty. So from a consumer's point of view, that's real world. That's what they live with. And that's what we measured.
At least with the miniSwap/U3, you choose the drive you want. And you can upgrade to faster and large capacity drive without buying another sealed unit.
Then please do title and run your report as a report comparing two consumer products, one USB3 and one Thunderbolt/SATA. Don't try to compare USB3 vs. Thunderbolt because that's not what you've done.
Thunderbolt, as used with external SATA drives, is a PCIe connection. If the implementation is good on the host side, it's the same as a PCIe slot on the motherboard. The important variable is the SATA controller chip in the external disk enclosure. Hold the SSD variable constant, compare multiple Thunderbolt enclosures, and pick the best one.
Then on the USB3 side: The controller on the HOST matters. Find the best motherboard implementation of USB3 out there (unless you are doing Mac-specific testing), and THEN vary the external enclosure, which is also containing a SATA controller (just like the Thunderbolt scenario).
Thunderbolt is as "native" or "transparent" as you get. It's simply an extension of the PCIe bus.
This doesn't mean USB3 is slow. But it cannot get faster than Thunderbolt if all other variables are kept constant. It's an additional protocol in between what are otherwise the same components in a system. There's no "magic" that makes data that traverses USB3 faster than if it skipped the whole USB3 translation in the first place.
----------
I am doing what I can to change the industry to measure SSDs in ways that actually matter... and provide useful information. I am in a position where I can affect this... but it is a very difficult problem, and one that will not be solved overnight. In the mean time... I just acknowledge that the current ways that we measure SSD performance are first order -- irrelevant. Such is life.
/Jim
IOPs are huge, bandwidth is important as well. That's why an old crusty 40GB Intel SSD feels faster than a hard drive, because it does IOPs better than the hard drive. It certainly doesn't do bandwidth better!
I work in the solid state industry, with a focus on enterprise (not desktop) class devices. These desktop storage devices are toys to me (but cool enough that I use them for my desktop). To see online publications oversimplify things just shows me that they have no idea what the real datacenter I/O traffic does to a storage subsystem. Where is the multithreaded testing? Where is the queue depth variance? What about the differences in the SSDs under test? That's the biggest single variable in this test, and it was utterly ignored.
