You are cherry-picking images and cameras that seem to support your belief, and perhaps the conversion style you are using squeezes some extra highlights in and compressing them in the top 2 or 3 percent of the display RGB values, like your bright cloud example, but one black swan, like a saturated red tulip converted to sRGB, brings your theory down, even with cameras with minimal raw headroom.
Yes, some subject matter, some cameras, at some ISOs, with some "picture styles" or conversion styles might show little difference in the headroom in a raw histogram vs a display RGB histogram, but that is not proof that there is never a big difference.
One can make a claim that "working people are paid according to their age, one thousand dollars per year of age", and then find examples for which this is incidentally true, which is probably made easier by the fact that people generally make more money as they age in the workforce, raising the rate of incidence, but the assertion is nonsense. The truth would be that on average, people's earning increase as they age, and sometimes, people are making their age in thousands per year and tracking close to it over a period of years. I probably had some years like that, but more that were not like that. Producing one year-end tax statement for which the claim seems to be true is not proof of the claim.
Accuracy for what? The histogram of a converted image is "spot on" for the converted image, but not "spot on" for determining how much more exposure you may have been able to use at that ISO setting in a raw capture. Would you prefer to shoot at ISO 40 on a camera that only goes down to "ISO 100"? You may be able to, but the histogram of a default converted image will never tell you if you can do it safely.
Which will be shown as 'white' on the final image. Objects that are 'white', that is pure matt white without specular reflection, will be shown as a light grey in a 'normally' processed image.
You have a wrong idea what tests are. What you are doing is un-systematic, partial, and ignores any comments you are offered. I have no idea why you are spending your time on that.
I don't offer opinions on images taken by other people.
Not to mention it's off-topic.
If you want to take better images, try to understand the difference between raw and processed.
It depends on where you place the threshold for "screwing up".
If you are happy with your images all well and good but you seem to be struggling with the fact that people have shown how image quality can be improved beyond what you are prepared to accept.
People have also shown how your approach to "testing" and the conclusions you draw from it are flawed.
i prefer Fast raw viewer. RD makes no sense to me at all. when i tune a guitar i use a simple guitar tuner . couldnt care less about harmonics or if the bridge is placed in the utimate position. thats why i buy good quality gear .let the manufacturer worry about that. its why i own an a74.