Here's most of what I know about EDFs in terms of design and flying, sorry for the wall of text.
A term known as "Fan Swept Area" refers to the frontal area of the fan disk minus the diameter of the spinner/hub. For example, an EDF with a diameter of 70mm and a hub diameter of 28mm has a fan swept area of
π( (70mm/2)² - (28mm/2)² ) = (Converting diameter to radius by dividing by 2, area of a circle = π * radius²)
π( 35²mm² - 14²mm²) = 3233 mm².
Thus, for this particular fan, an exhaust duct with a cross-sectional outlet area of 3233 mm² is said to be "100% FSA": The area of the exhaust is 100% that of the Fan Swept Area. An outlet of 90% FSA is (.9 * 3233 mm²) = 2910 mm² and an outlet of 105% is (1.05 * 3233 mm²) = 3395 mm², again, for this particular example EDF. For a circular exhaust area, the radius can be found by backsolving A = πr².
The purpose of the exhaust duct (or "thrust tube") is to get the flow stabilized and accelerate the flow, providing a higher "efflux" (exit) velocity than the bare EDF. Changing the outlet FSA changes efflux velocity and also changes static thrust. Generally, keeping the outlet area at 100% FSA will provide maximum static (on-the-ground, not moving) thrust at the expense of efflux velocity. Constricting the outlet area increases efflux but reduces static thrust, which can result in quick jets which take a bit of doing to get off the ground. There's a certain balance that needs to be achieved here if one is optimising for speed, but 100% FSA is a good place to start and is what I use personally - Going fast isn't as big of a concern for me as having better low-speed throttle characteristics, which can be useful during takeoff, landing, and advertent hairy situations. If you'd like to go faster, around 85% is a reasonable lower limit for most EDF units (cheaper ones reportedly perform better around 90% or higher). It is important to note that not all EDF units and not all airframes are the same, so the optimal outlet for your specific scenario can only really be found through trial and error.
Thrust tube geometry is also somewhat important. I've got the best static thrust results with a thrust tube that is cylindrical up until the end of the motor bell and then tapers linearly to the exit from there, the idea being the air isn't needlessly constricted in the annulus between the motor bell and wall of the thrust tube as would be true in a constant taper design. The performance difference is relatively small (<5% static thrust), but it is there. Thrust tubes also give the EDF a nice "woosh" sound, with the effect being more pronounced the longer the tube is.
Intake ducting is some of the same but with caveats. I personally recommend having an inlet FSA of a tad over 100% for the purposes of countering effective losses in area from the boundary layer. Constricting the inlet area reduces drag and shifts the operating range of the EDF towards higher speeds, much like constricting the exhaust outlet area does.
Inlet geometry is also important. try to keep the inlet ducts at the same cross-sectional area as the inlet itself throughout their length (maybe a tad larger to counteract boundary layer) and try to provide a smooth curve into the EDF. Avoid sharp corners or edges on the lip of the inlet and internally, as they can cause flow seperation and turbulence which reduce performance.
Generally, planes with EDFs are much more inefficent in propulsion than their prop counterparts. This necessitates big batteries and high wing loading for reasonable flight times, which leads to some very quick birds. EDFs are also different from prop planes in that they have no prop blast over the tail surfaces. If you're used to "punching out" of a bad situation with throttle or dabble heavily in 3D, keep in mind that there's no easy way to increase control authority other than speed - and getting speed is different too. Prop planes are far lower pitched than EDFs and as such have relatively immediate throttle response. EDFs have to deal with a high pitch and the inertia of the air moving into the inlet duct, through the fan, and out of the thrust tube. As such, the RPM response of EDFs is usually quite quick, but the thrust response isn't immediately there, which requires you to "fly ahead of your airplane".
In RC, we typically think of using power to control airspeed and elevator to control altitude. With an EDF jet coming in on final, in high alpha, on the back side of the power curve, this doesn't work too well: think of Elevator as controlling the flight speed, throttle as controlling the altitude. Also keep power in on final, don't just chop the throttle and bring it in with the elevator. EDFs demand you to carry some power for good landings, which also helps eliminate some of the thrust lag time when some more power is needed.
Things happen a lot faster with EDFs than they do with props. They're a lot more stressful on their gear and their pilots but can be great fun once you've got comfortable with them.