Here's another DR comparison where the only change is replacing 12mp DR 100
with 12mp DR 400, which is
obviously using a different tone curve to increase the DR,
not EXR hardware that would provide a "true" DR increase.
No one picked you up on this major boo-boo so it's left to me. ;-(
You didn't even look at your own image - in the bottom part all tone curves are identical.
In the top part (which you left out) DR200 curves, whether Software or Hardware, are superimposed over each other.
It's the same for DR400 - both Software and Hardware tone curves are identical.
There is no such thing as a non-tone-curve-related "true" DR increase. In other words, what you call "true" DR and others call EXR Hardware, is still totally tone curve limited. It's exactly the same as good old EXR Software...
Someone keeps flashing up these DPR charts to try to show how I have made a big mistake, but here is the next chart down from the one you have shown. I have added annotations:

Please read the text below the chart - it's a very good description of Hardware versus Software EXR results
There is no "magical" increase of Hardware DR over Software DR, nor is there a big noise problem at the bottom end.
In fact, DPR's chart (where I wouldn't be surprised if they had accidentally reversed the proper IDs) shows a slight 2/3EV improvement in Software DR in the lowlights. Wow.
What's more, my independant testing, without even realising that the results were home and hosed already, shows the same thing but a little more visually. It's no wonder, then, that the later X-Trans sensor in the X20 produces a slightly better result than from the X10.
I don't read your posts normally, but I was drawn to this one!
--
Cheers ;-)
Trevor G
Silkypix tutorials at:
http://photo.computerwyse.com