Below is a screengrab in Photoshop off my calibrated sRGB-ish Dell 2209WA LCD of a 100% cropped view of an image of a decorative deep blue glass crystal ball captured in Raw under outdoor daylight shade and processed in 16bit ProPhotoRGB ACR to demonstrate why sRGB sucks as a technical quantitative comparator and even a working space no matter how many perceptual levels of colors it encompasses.
You may be correct, but your demonstration is a bit lacking in detail. This is what happened in the Unconverted image:
1) opened a raw file in humongous ProPhoto
2) automatic/manual adjustments were applied in the raw converter spinning colors around
3) color management then tried to squeeze these wide ranging ProPhoto colors into your monitor's presumed color space using unspecified parameters
4) the video card driver and LUT performed more squeezing - a screen capture was taken
5) the monitor displayed its best (sRGBish) rendition of such corrected colors
In the Converted image you went back to step 2) and inserted step 2a) after it in the list:
2a) conversion to sRGB color space using unspecified parameters.
The two screen shots were compared, but there are many unknowns for the comparison to be meaningful.
Since in 99% of cases people are either printing or viewing in sRGB, it would be interesting to know whether it makes a
practical difference in day to day use to work in it
from the very start or rather convert to it at the end of PP. Squeeze at the beginning or at the end, but always squeeze we (non-pros) must. My feeling is that most of the color differences/shifts are introduced during conversion into the final color space (sRGB), and that the more and the more extreme adjustments one makes in a larger color space,
the more the chances that final colors will be wild guesstimates - which may then need to be corrected for (again if converted later) after conversion. On the other hand when there are minimal adjustments, there is no reason why Camera Space --> XYZ --> ProPhoto --> XYZ --> sRGB at the very start should not result in very similar values to Camera Space --> XYZ --> sRGB, assuming proper conversion parameters. If I understand correctly this is what is shown in your attachment (the conversion to sRGBish in the left image capture performed by your CM system). So where do the differences come from? The first suspect that comes to mind is the chosen intent/compensation, the second are adjustments performed in ProPhoto. BTW, the more direct approach should in theory result in less noise.
To check this with your difficult test file, you could open the Raw file directly into sRGB with an sRGBV4 profile (Nikon's is good) using a variety of intents with/without blackpoint compensation and produce a series of comparisons to ProPhoto+sRGB in the same conditions. Better yet you could post the Raw file and let us have a go at it with our own raw converters and CM workflows.
Jack
[EDIT]
I viewed this page in Chrome, which is not color managed. Here is a screen capture of tlooknbills' message and thumbnail in Chrome, superimposed with the relative file open in color-managed CS5. Lost some of its purpliness already. I wonder what it would look like if originally opened in sRGB direct from Raw by a well behaved Raw converter other than LR.