I got called "full of it" and other nasty things for expressing my skepticism of Thunderbolt and its SATA controllers (and moreover, was stating I was seriously worried Apple might take all our internal drive bays away... looks like my fears were well-founded on that one). My concern is that it was a serious downgrade from having internal "On board" SATA in terms of latency. I was also told that any limitation found in the benchmarks was clearly the fault of an inferior Thunderbolt SATA controller, and that I needed to find benchmarks where, somehow, the Same chipset was used for thunderbolt as PCIe... My opinion is that, as a practical matter, if there exist no Tbolt controllers that are as good as onboard SATA, TBolt is is a downgrade. I just want to know either way, as I can't find any real-world evidence either way. The only reviews I can find benchmarking TB's SATA controllers either compare it to USB 3 (which, duh, it beats by a fair margin), or losing horribly to SAS on sequential read. I was told that this was simply because the SAS card was 8 lane and TB 1 is the equivalent of 4... in spite of the fact that TBolt supposedly tops out at 1GBps. What I'm really more interested is random-write/random read latency. I have had a lot of experience with Firewire bridgeboards and was thoroughly annoyed with "800mbps" being a bigger lie than the cake, not to mention the latency which was noticeable even with just opening a single small file. I am not a fan of TB as a replacement for onboard SATA--I much prefer putting all my hard drives INSIDE the case and only having to pay for 1 case and 1 PSU to power my computer. However, I do acknowledge that eventually, this technology will become cheaper and external options will stop being utterly insane. If TB can replace SATA without issues, I say the more interfaces, the merrier!