Unless they post performance specs rated for sata 2, then you simply just don't know if these newer drives are any faster than the current bunch at least as far as on a mac pro is concerned. And I'm not comparing the corsair's marvel chipset with Intel, I'm just saying that a high sata 3 rating doesn't do us much good at the moment. Also the random speeds on the corsair P3 were worse than the random speeds on the vertex 2. Hopefully, that is just on the Corsair's sata 3 drive and won't be a trend.
As per SATA II and SATA III, the difference is the bandwidth of the specification (SATA III doubles that of SATA II). There can be differences in controllers with the same specs (i.e. AMD v. Intel, Marvell, ... located on the system board, not the drive), but it's more on the drive controller than the ICH used in Intel based systems (Intel is the only company that uses their own design that I'm aware of; others are buying parts from 3rd party companies such as SandForce or Samesung... err... Samsung for example
).
The specifications as to how the data is transferred are the same (why the controllers can be made backwards compatible in the first place - all they do is throttle the available bandwidth to the slower specification used in the chain). So there will be cases where a SATA II controller will reduce the sequential throughputs as it won't have sufficient bandwidth to handle it (cases where the sequential speed of SATA III based disks can exceed ~2375MB/s, which is the real world limit for SATA II). Burst speeds too, but they're not as important as sustained sequential and random access (even SATA II should be sufficient to handle the Emcrest disks for random access if it does turn out faster than Postville's, which it should once they get the drive's firmware finalized).
BTW, ~70MB/s is all the fastest SSD's currently available can do. So even if this doubles, the SATA II based ICH10 used in the MP can handle it for random access. It's sequential access that the ICH10 will throttle, as the current disks run at the limit of SATA II. It will still work due to backwards compatability in this instance, but it won't run as fast as the drive would be capable of (applicable to most any SATA III SSD that's to release in the near future - the budget models would be the potential exception, such as the Value line from Intel).
I think this is what
Transportuer was trying to explain.