Yes, according to Thomas Knoll:
Internally, the S&R images are first merged into a single HDR image. The S pixels are used for lower part of the tone range, and the R pixel are used top of the range, with a smooth transition between the two. This merging is completely independent of all the camera raw controls. The exposure and recovery sliders adjust the white clipping point in this merged HDR image. The other tone controls adjust the rendering of the values between zero and the white clipping point.
In the same thread it was also asked if the S&R pixels were handled under Compressed RAF and DNG formats, to which Thomas replied:
Yes and Yes (both).
For the full thread, see this Adobe thread.