How the hell did I miss this thread?
Couple things -
A) Compressor wheel flow is related directly to the housing it is in. The CT compressor housing is small, and a 67mm compressor wheel (I'll assume you are talking the 67mm T04R aka T67 wheel) with an 84mm exducer. That's about 2.64"/3.31". The housing inlet is only about 2.75" - that is a very thin wall to be clamping on. 3.31" is rather large to be bolting into that housing - your flow will suffer greatly, partly because of my next point -
B) The turbine is freaking tiny. Sure, you may flow better in the mid range, and balancing issues aside (It's not the mass difference you are worried about - a large wheel spinning at xxx,xxx rpm will amplify ANY imperfection in the balancing, vs a small wheel spinning the same rpm)
C) Spinning that tiny turbine full blast, at least say... 120,000rpm. The compressor wheel is also spinning at 120,000rpm. Big deal, you say? Well sure, the angular velocity is equivalent, but guess what - the linear velocity IS NOT.
That being said - Trigonometry to the rescue!
Let's take our turbine, spinning 120,000rpm. That's 2000 rotations/second. What's the angle measurement of any circle? 2pi radians = 360 degrees. So we're spinning 4000pi radians/second. Moving pretty good!
Ok, so for the turbine, what is our inducer size? 2.36", is it not? Linear velocity = radius of the circle * angular velocity. So 1.18 inches * 4000pi rad/sec = 4720pi inches/second ~ 14850 inches/second. Check the math, but I believe that is about 843mph. What does that tell us? That the limit of the CT26 turbine housing is a hell of a lot less that 120,000 rpm - You should never exceed the speed of sound at the tips of your wheels! The speed of sound is ~761 mph, so I would expect the limit of the CT26 turbine speed to be approx ~108,000 rpm, which would equal ~760mph.
Head hurt yet? Ok, now lets remember that the compressor is going to spin JUST AS FAST as the turbine. 1.655" radius on that 84mm exducer! Same calculations now gives us a whopping linear velocity of 1065 mph! Think that's going to flow? Not just no, but hell no!
Flip side of this pretty mathematical picture is that the compressor wheel is going to massively increase drag and loose flow when the tips hit the sound barrier. Somewhere just north of 77,000 rpm. (which is again, 760mph).
This means the exhaust turbine now fights to spin the compressor wheel faster - further choking the exhaust side of this equation.
This is the point where I severely wish I had a flow map of the CT26 turbine!
Anywho, quick comparo to the standard 57 trim (which already gets a boost, as it will better flow with the housing) Runs a 2.95" exducer, so you get a maximum speed of ~87000 rpm before the turbine starts to fight it. Extra 10,000 rpm of that tiny exhaust turbine, coupled with better conformity to the flow characteristics of the compressor housing, equals win.