"A five stop range of reflectance is assumed..."
Why?
The assumption is based on an observation and a deduction. The observation is that when the illumination is uniform on a subject the maximum subject luminance range is rarely greater than 5 stops.
The deduction is that any object that is evenly illuminated (no glare, no shadows) will fit into a 5-stop range, which is 2 stops less than the normal 7 stops.
Davis writes, "it seems logical to assume that the subject luminance range can be estimated by adding the illuminace range, as measured with an incident meter, to the basic subject reflectance range of 5 stops. In practice this works surprisingly well."
In practice the illuminance range is measured by taking two incident readings, one in the shadows where you expect to have some texture, another in the most brightly illuminated area of the scene. You add this value to 5 and you get the subject brightness range of the scene, as it is understood in BTZS terms.
BTW, assumptions are not unique to BTZS and incident systems of metering. ZS is also based on a fair number of assumptions.
I have been using ZS virtually all of my adult life, and BTZS since about 1990. Both methods have their advantages and disadvantages, but on the whole I find BTZS more accurate because it is based on testing procedures that are both more extensive and more precise than the traditional testing procedures of ZS. However, the real beauty of BTZS is that results of the testing, once done, can be used as the basis for using either ZS (reflectance) type metering or BTZS (incident) type metering in real life situations. And this is very useful because there are definitely conditions that are better suited to one or the other of the two systems. For the most part, however, neither system is any more complicated or simple than the other, and regardless of which system one choses to use, the most important decisions we have to make are still creative decisions based on how we want our tonal values to look on the final print.
Sandy King