You've got the basics of fly sparging, but it helps to understand some of the gotchas.
With batch sparging, you start by vorlaufing, and draining the first runnings. The gravity of these runnings should be pretty high, and will remain more or less constant during the drain. Then you add a "batch" of sparge water, stir, vorlauf, and drain again. The gravity of these runnings will be lower than the first runnings, but again will remain fairly constant. If necessary, add another batch, stir, vorlauf, and drain again etc. The gravity of the runnings from each batch will be lower than the previous.
With fly sparging, you vorlauf once before starting the sparge, and then drain the wort out of the lauter tun while adding sparge water at the same rate as it is draining. In this case, the gravity of the runnings decreases as the sparge progresses. The gravity decreases slowly at the start of the sparge, but more rapidly as the sparge progresses.
The big difference between the two methods is the way in which the sugars are dissolved into the wort.
With batch sparging, the sugars are dissolved by the stirring, while with fly sparging, the sugars are dissolved by the sparge water gently flowing through the grain bed.
For fly sparging to work efficiently, it is necessary that the sparge water flows evenly through the entire grain bed so that it can rinse the sugars out of all the grain. This means that the sparge water has to be delivered across the entire surface area of the grain. I use a rotating sparge arm which works well. I think ODaniel's solution would work just as well, and save a few bucks. There are also other methods that would work, and the purpose of keeping a couple inches of sparge water above the grain bed is simply to prevent the sparge water delivery mechanism from drilling holes in the top of the grain bed, and providing an easy path for the sparge water to reach the bottom. This is called channeling, and if it happens can severely reduce your efficiency.
The next thing to consider is the collection mechanism at the bottom of the lauter tun. The best device is a false bottom that collects wort evenly across the whole of the bottom of the tun. Next best is a manifold with cross bars that doesn't leave any substantial gaps where wort will not be collected from. The worst method of collection is a braid, but even these can work with fly sparging if they are long enough, and meander around the bottom of the tun to provide a good collection area.
Then, you have to pay attention to the flow rate. If you add and drain water too quickly, it will find the easiest way from the top to the bottom, resulting in channeling, and leaving some of the grain unrinsed. In the on line version of How to Brew, Palmer says 1 qt per minute is most common.
In the book, he states that 1 qt per minute is the maximum. In my experience, the flow rate is dependent on the collection area. (A larger area allows a faster flow) With my 5g cooler MLT with a false bottom, I cannot exceed 1 pint per minute without getting channeling. With my 10g cooler MLT, I can get up to 1.75 pints per minute (possibly more).
Finally, I can get about 5% better efficiency with fly sparging than with batch (with an OG of ~1.050). If you don't have a setup that is optimal for fly sparging, you may well get better efficiency in less time with a batch sparge. On the other hand, I get the same efficiency when fly sparging with high gravity brews (~1.080) although I have to sparge slower. Although I have never tried it, I believe that efficiency decreases when batch sparging with higher gravity brews.
Hope this helps
-a.