I know this link since the past milenium, this is nothing new for me.
I don't know Poser, but even my eight years old rendering engine the colors are in floating point, in the rastering process are converted to fixed 16 bits per color (RGBA), but the final result is 16 bit RGB (not 24 bit). Last week I reworked my engine after many years of pause to give 24 bit RGB.
But you have nothing to do as the final media do not support higher dynamic range. If you compress the range the result will a distorted image. If you take the higher part the result will be that most of the scene is black (happens with photograph). If you truncate the image is good with the exception of uggly saturated white spots (happens with Poser or digital cameras).
With special high intensity monitors the bright part will hurt your eyes.
What I know is that the research of the guy of this link is something different, he use HDRI images for environmental mapping. You take a box and project six images of a cathedral in each face of the cube, your camera is inside the cube, so your scenario is the cathedral. Normal images have not enough dynamic range for all the different illuminations in each place, that's the reason of HDRI, but once you focused to camera in one part of the scenario you return to the old 24 bit RGB and the result is normal.
Of course if your camera is fixed you can put a normal image as the background, but if you want to have a 360 degree movable camera you need HDRI images mapped in a box or sphere.
That is why I asked what people are trying to say what is HRDI.

Stupidity also evolves!