Thats right Heinz! not so much the design (I am not that good a craftsman) but the concept anyway.
I guess it depends on how much water you need to process, myself I have about 1200 so I am stuck with a large skimmer no matter what. Prior to making mine I did go shopping for something to handle my tank (and the way I run my tank) I had David pound me pretty hard to get one of his deltecs, but man I couldnt justify 4 to 5K. Mine ended up costing me around 600 bucks.
Width of the mixing chamber add to the compacity of the skimmer, which in turn adds to the skimmer dwell time for the water moles. Skimming in general is two basic princables. One is mechanical striping of the organics attached to the water moles. This is done by the air bubbles pounding into the water moles and stripping the organics that are attached to it. The more air bubbles the water mole has to pass the more of a beating it gets. This process is refered to as the bimbardment rate. The second process is the chemical attachment of the free organic to the surface of a clean air bubble surface. This process has to things you need to work well, one is as many clean air bubbles as you can have (usually about 20% of total water volume of the skimmer) so that thier is always surfaces available. The second is the right ammount of dwell time for both the air bubble and the water mole.
Organics just dont jump on the air bubble, most could never attach. What happens is the first organic to attach (surficants which live at the water/air interface) to the bubble, this changes the air bubble surface from clean to a surface covered with attached surficants, from here many more organics can attach themselves to the surficants that now cover the air bubbles. So this process takes a bit of time, thus making dwell time so important.
Anyway sorry for the ramble, we did a great thread on the process and the math behind it in Anthony's forum on RF, its called Skimming 101.
mike