I have extensively tested the CAI and the POP on the dyno and have
  found that CAI produces more power than the POP alone from 3500 rpm to the fuel cut.
    The gains were 8 hp@5500 rpm and 3 hp at the power peak. Past the power peak the
  CAI continued to produce more power.  The POP was good for about 2-3 hp over stock on
  a mildly modified motor. 
  I wrote an article in the March issue of SCC about the CAI among
  other things that has dyno charts and stuff so you can see for yourself. 
  Clark and Jim at JWT have some valid concerns. In their experience
  Intakes with non-radiuses inlets can cause turbulence that causes irregular MAF voltage
  readings that cause driveabilty problems particularly at low rpm. 
  The set up that I and Mike Mager use has an adapter to use the JWT
  POP with the CAI.  This version is available for sale and I believe that this is the
  one to get.  It has the advantages of both the POP(bigger filter element, radiuses,
  non-turbulent base) and the CAI (cold air and resonate supercharging). 
  Justin Choi's statement that the main power increase that the CAI
  has is due to the length rather than just cold air is correct.  The added length
  contributes to resonance that increases VE. 
  Both Justin (MAF voltage), Mike (1/4 mile time), and myself (dyno)
  have assembled repeatable quantitative data that is not based on seat of the pants
  guessing. 
  This should close the debate unless someone else comes up with
  repeatable data to the contrary.