I did say I was going to attempt the impossible . I can see now much more clearly why reviewers like Dpreview only test in-camera jpegs. The differences between RAW converters and the possibility that some might apply a default sharpening that can't be turned off, would make the test results invalid.
I did some more conversions of the previous image, but this time in RSP I turned sharpening to it lowest point, minus 50, and 'detail extraction' to its lowest point of minus 50.
The first set of 3 crops includes 'local contrast enhancement' to the ACR image, as someone suggested. Perhaps a pixel radius of 50 is too great, hower the 'unsharpened' RSP image with its minimum settings still looks the best.
In the second set of 3 crops, I've applied Photoshop's Smart Sharpen in default mode. This brings out the detail in the ACR image to a level that is comparable to the unsharpened RSP image, but what happens if I now apply a bit of sharpening to the RSP image?
The fact is, one could spend a lot of time messing around with various sharpening methods and degrees of adjustment, which is why I prefer time-saving automatic features.
Perhaps the bottom line is, RSP seems to deliver better results 'out of the box', so to speak.
One other point I find interesting is that these extreme blow-ups of very small crops, show the same detail and subtle shading when viewing the enlarged thumbnail from within LL, as they do when I open the same jpeg compressed image in Photoshop.
This fact seems to be at odds with some hints I read from some quarters on the Rob Galbraith forum recently, that the web is a very poor medium for displaying subtle differences between images.