In the last months I moved from ACR to DCRAW. A very different philosophy: I use DCRAW to get a high quality plain linear image with no adjustment applied at all, and with total control over how the development was done. In difficult or important images I can analyse in detail the histogram of this linear image to find out EXACTLY how many pixles, in which channels and in which areas are blown and test different hightlight recovery strategies if needed.
Then I transfer to PS all the non pure developing tasks: exposure correction, camera profile assignment (still in linear mode), colour profile conversion with gamma, contrast, saturation, noise reduction, sharpening,...
The only thing I miss from ACR is the colour noise reduction, which performed very well there. The rest is unnecesary for me and PS can do much better (for instance being fond of curves, ACR curves alter Hue while PS curves in Luminance blending mode are perfect preserving Hue).
I know this will be never a popular way to proceed, but for me is fine. I call DCRAW from many of my programs so I am used to deal with linear images that have some advantages. I like the concept!
