The link I posted is hundreds of tests of every sluice design possible at the time. The conclusions which they came to are what I based the 8 inch max size around.
Here's some snippets from their conclusions on box width and why its better to use an 8 inch width for river/stream powered equipment. For a high-banker, its ok to go somewhat wider. You can read in detail what they found out about high-bankers somewhere in there as well. The end of page 13 (volume III) says:
----------------------------
Sluice Sizes
I have mentioned that as I was building my original sluice as described in Part-I of this article I began to have second thoughts about it’s size but after using it I was glad I stuck with the 8-inch width and didn’t make it wider. After running these tests I am even more satisfied with using an 8-inch wide sluice as compared to a wider box.
Friends told me that I was hindering myself since I can’t run a lot of material in a narrow box but based on what we experienced this summer we came to the conclusion that sluice width is not one of the critical factors in performance or processing capacity, up to a point, and that is where one switches from a stream sluice to a power-sluice.
Sluice Building
As long as a person is shoveling material into a sluice of any type an 8-inch wide box readily accepts a shovel load of material just as efficiently as a 12-inch wide box but actually washes and stratifies that material much more effectively. In fact I’d go so far as to say that for a manually operated sluice the narrow box outperforms a wider box in almost all respects.
An 8-inch box not only easily accepts a full shovel load of material but it uses almost 25% less water volume to process that same material load compared to a wider box if we look at water velocity between various box widths. For a stream sluice this is an important consideration since we’re usually very limited from a velocity standpoint and velocity is what actually gets a set of riffles operating properly. It is water velocity that actually does the work in a sluice far more so than water volume and narrower boxes naturally have higher velocity than a wider box given identical water volumes.
Narrower boxes also have the advantage in that they can be made much shorter than wider boxes due to the effects of material distribution as described in the following section.
-------------------------------
Here's another section on the bottom of page 18 (volume III) which explains why narrower is better:
-------------------------------
The ‘capture’ area of a riffled sluice can actually be fairly short in comparison to a sluice that only uses mesh over moss or carpeting. From evaluating other studies the riffled capture area of boxes that work best seems to be around one square foot per every 100 gallons of total water flow per minute. In other words for a box designed to handle 100gpm of flow a riffle section containing 1 square foot of total area is near optimum. Using this rule of thumb usually results in a length long enough for a run of 6 to 7 riffles after the slick-plate and will give you around 96-square inches of net clear deposition bed in an 8-inch wide stream sluice. For a power sluice this length and the number of riffles can be increased by 25% for better results at higher flow velocities. [AstroBouncer's note: the power sluice he is referring to here is a highbanker, not a sluice]
It is tempting to build sluices wider in an attempt to increase the area of the riffled collection area without making the sluice longer but keep in mind that adding only 2-inches to the width of a box increases the collection area by only 20% but will require the use of 50% more water flow to maintain the same velocity and it is the velocity that provides the energy to drive the riffle vortexes. For this reason it is always better to make the box longer if more collection area is needed and not wider.
-------------------------------------
And lastly from the testing conclusions, on page 62 volume III:
-------------------------------------
3. Flow velocity is far more important than flow volume and even sluices designed exclusively for fine gold recovery can benefit from high flow velocities. In sluices using riffles it is the velocity that provides the energy needed to get riffles of any type up to their ‘working’ level.
4. Narrower sluices are more ‘efficient’ than wider sluices for any given length so if in doubt go narrower and longer rather than wider and shorter. It is much easier to have better control of velocity in narrow boxes and recovery of both large and small pay materials depend on velocity and not volume
-----------------------------------
Before I read that article based on scientific testing and data collecting, I always thought wider was better too. I don't think anyone has ever done more testing on different sluice designs anywhere, or if they have they haven't documented it like that and enabled the public to read it. I should note also that there is a whole lot more information as to why to use 8 inches rather then a larger size, but as that is rather wordy and deals with different velocities, mathematics, and how much dirt a single person can move, I decided not to include that information. You can read about it in their testing though at the links I provided.
EDIT: I was thinking about this some more, and one more thing I want to say is, all of the above only applies to the testing done with conventional riffles, IE Angled iron, lazy Z, or Hungarian style, etc.None of that applies to the newer technology of drop riffle sluices such as the Angus McKirk, Le Trap, California sluice box and more. For those sluices the flow is much different because the vortexes used to swirl the material are under the main flow of water rather then in them. So a 12 inch sluice box that is a drop riffle type such as a Le Trap, is perfectly fine for one person.
What type of sluice box are you making?
You can see some of my custom boxes here:
My first sluice (sluice in the middle):
Newer picture of my sluice boxes, the tall one is my main sluice and its eight inches wide. The small one next to the carpet sluice is a prototype using a new drop riffle design that has been working well in testing.
Drop riffle before the angle cuts :
Drop riffle hybrid after the new cuts (so it cleared lighter material faster):
Another thing to think about is how wide you can make it and still fit it in a 5 gallon bucket for easy cleanups. The inside of a 5 gallon bucket is about 11 and one quarter inches wide.