Hello, I would like to clarify the following. If in ACR 2 (CS) you open a 16bit image, work on it in ACR and then open it into PS CS with the Depth set at '8 Bits/Channel' does ACR perform all your ACR corrections in 16 bit mode and then convert it to 8 bits for opening in PS CS, thus preserving as much data as possible?
I am sure (and would hope) that it does.
Since ,in that version of ACR, you can't work with tiff or jpeg, you are working with the full raw file. It won't get converted to 8 bits until you open it in PS.