the way a fluidized bed filter works is pretty simple, and similar to how any other bio-filter works. the only key difference is the bio- material is kept suspended in a moving water collum. instead of using bioballs (or floss, inert ceramic pieces,etc) sand-like material is used. the sandy material provides a load of surface area for bacteria, and moving water keeps it suspended and moving through the water collum in a 'sealed' container. the suspension keeps the bacteria with a never ending diet of dissolved chems like ammonia and nitrites. through the rapid water movement it's actually as efficient or more so than undergravel filters, bio balls in a sump, etc.
the only problem with fluidized bed filters, is getting the right amount of fluidization, too much and the sand will be held at the top of the container where it has less time to absorb any of the dangerous chems, and sometimes be discharged out of the system, or too little and the sand clumps together and sits on the bottom of the container doing little to nothing for you. the most efficient 'system' holds the sand in the middle of the container, constantly being mix with new tank water, and water about to leave the filter.
as for the diy
page, i think it's way to much work and the page is not detailed enough to attempt (txt is fine, but a picture is worth a thousand words). and i don't like the fact that construction is mostly trial and error, 'this has to be 3/8 exactly or it's not going to work' statements have me a bit worried, and the process of priming the system seems very work intensive, this is a hobby afterall isn't it?
a good fluidized bed filter can be had for 20-50 bucks [www.thatpetplace.com]
and comes with everything but a pump. most of the new designs have self-adjusting fluidization, where it will limit incoming water to reach the best efficienty. sometimes it's better to re-invent the wheel, then others it's better to just use something off the shelf.