The plenum size is dependent on overall airflow as well as boost pressure (to the extent that it affects mass airflow).
Your overall airflow is dependent on displacement, your *true* volumetric efficiency (dependent on head flow, pressure drop, cam overlap, etc), not your "turbocharged" VE, which is greater than 1.
The runner length is dependent on where you want the "hotspots" to be. In other words, just like headers, the length is dependent on a pseudo standing wave being present at a resonant frequency that corresponds to valve opening events. (Once per 2 revolutions, divide by 60 to convert to Hz from RPM)
The idea for the intake is that the plenum is maintaining a constant, or near constant air pressure at a given flow rate. Naturally, this means the one you want is larger in volume than the stock one. Big enough that the pressure remains fairly constant, small enough that your turbo can actually keep it pressurized. I honestly don't know the math off the top of my head for that calculation, but that's the basic idea.
On the other end, you want to have another high pressure point right outside the intake valves. Since this requires a standing wave, you have to "tune" it for specific RPM ranges
The formula for resonant frequency is pretty simple at first glance. Frequency in Hz = V/2L.
L = length of runners.
V = Speed of Sound in the medium. 331 m/s at 21 degrees C. (Temperature is delta from 21 C)
V = 331.3 + (.606 * T)
The last part is where it gets tricky. Luckily, the speed of sound is almost independent of air pressure, which seems counterintuitive when you're talking about setting up pressure waves. It's best, unless you want to get really really hardcore, to assume that air is an ideal gas.
Even if you assume an ideal gas, you still need to start looking at compressor maps and intercooler efficiency and pressure drop, so that you can arrive at:
A) The actual output pressure of the turbo (your boost pressure + the pressure drop of your intercooler system), which you can then use to find:
B) Your intake charge temperature, based on the pressure ratio and efficiency that your turbo is running at.
From there, you can then calculate the speed of sound in your intake, and then choose your target rpm range. If you get fancy, you can try and tune it for two different frequencies for both for off-boost (ambient temperature) and full boost (intake charge temp), since the resonant frequency is increased at higher temperature.
It is also important to remember that the speed of the pressure/sound waves through the runners is *independent* of the actual speed of the air moving through it. Also, humidity has a slight effect of increasing the speed of sound in the medium.
Now, I'll give you a little hint. Similar to header design, you're dealing with frequencies of 75 Hz or less. You'll soon realize that the tube length required for an ideal runner is much longer than will fit in your car. This is where we exploit the fact that waves can reflect, even in open ended tubes. In other words, the wave will run up and down the runner multiple times between each valve opening. The open end in the plenum effectively increases the tube length by .6*radius. Also, include the length from the head to the intake valve in the tube length, and assume it is closed (It should actually correspond to the high pressure wave peaking at maximum valve lift, but good luck doing the calculation and having it be accurate unless you want to start doing non-ideal gas behavior and airflow around the valve, etc. I guess you could add your valve lift to the length of the very last wave arrival too, but it doesn't seem reliable)
Anyway, you'll want to cut your ideal wave in an odd number of segments to get your runner length, so that the high pressure wave arrives at the valve at the right time. Since it isn't a true standing wave, the effect will not be as pronounced as a full length tube would be, but it has the added side effect that any wave that passes through the tube exactly an odd number of times will have a similar effect.
As far as runner diameter goes, it's basically a tradeoff between air velocity and pressure drop. The simplest way is to keep it the same cross sectional area as the intake valves.
Other variables to think about are having a slightly tapered plenum to keep the pressure equalized between the four runners, adding tapered "horns" to the runners inside the plenum to produce a ram-air effect (increases the effective pressure at the runner mouth).
These formulas are identical to the ones you'd use to calculate exhaust runner length for a manifold. The only difference is that you're tuning for the low pressure standing wave to promote scavenging between cylinders rather than using the plenum as positive pressure. Substitute EGT for intake charge temperature and you're there, assuming you use the exhaust valve diameter. The only funky part about exhaust is the voodoo involved in collector diameters, but since you're doing a turbo manifold instead of a N/A one, you can probably just taper to your turbo inlet.
Good luck!
There are some good books out there on turbocharged engine design:
Winterbone and Pearson; Design Techniques for Engine Manifolds, Wave Action Methods for IC Engines; Society of Automotive Engineers, Inc; 1999; ISBN 0-7680-0482-9
That book is quite expensive, which is why I don't own it yet.
Also, a book that's OOP:
Watson and Janota; Turbocharging the Internal Combustion Engine; The Macmillan Press Ltd; 1982; ISBN 0 333 24290 4