Holy shit batman, where to start...
First with some airflow numbers and spool times (boost will be 20 or 25 psi depending on the turbo)
14b - 30 lbs/min - 3000
small/big 16g - 40 lbs/min - 3300
EVO3 16g - 42-44 lbs/min - 3300
20g - 48 lbs/min - 3700 (tdo6H, tdo5h will be ~3500)
50 trim - 50-51 lbs/min - 3900
60-1 - 60-62 lbs/min - 4200-4400 (depends on hotside)
56 trim GT40 (GT35R) - 65 lbs/min 4400
T67 - 75 lbs/min - nearly 5000 on a 2.4
These are all measurements taken on my own car FYI.
In the case of the 16g vs 20 and movingmore airflow at the same boost pressure. Unlikely. The turbine sides are the same in the case of the TDO5 20g. The difference in compressor efficiency is almost impercievable in the real world. If the 20g is a TDO6H 20g, then you will get more airflow at the same boost thanks to the increase in turbine flow/volumetric efficiency. I have not done that comparison however. But to illustrate the point, when I ran a tdo6H 20g and a TDO6H 56 trim at 25 psi, airflow was exactly 48 lbs/min in both cases. No gain at all, despite a much larger and very efficient compressor wheel
The key to making power, especially on pump gas, is to move more air without raising boost. Airflow is power, boost is heat and heat is knock (laws of adiabatic compression/ideal gas law).
Some good examples of increasing airflow without raising boost, in addition to the larger turbine side, are cams and intake manifold. At 20 psi on a normal setup cams are worth 3-4 lbs/min. An intake manifold is typically worth ~3 lbs/min. This is all free power as far as pump gas goes, assuming the turbo has the extra flow capacity to meet these numbers. A camed car should move 40 lbs/min with cams at 20 psi. ~36 without cams. 1 psi is roughly 1 lb/min on a 2 liter, so thats the difference between getting 40 lbs/min at 20 psi or 24 psi.
Good intercooling is the other half of the pump gas puzzle.
Airflow is worth roughly 10 hp per lb/min. I like to size the turbo based on the airflow/boost I intend to run. For example, a 2 liter at 30 psi and 8000 rpm is only going to move ~55 lbs/min at 100% Volumetric Efficiency, so a 65 lb turbo in this case is not going to help a whole lot.
But that 65 lb wheel will be maxed out at 30 psi on a strocker. so it depends on the goals and setup. Going from the 20g to the 56 trim at the same 25 psi actually made the car slower because it was moving the same amount of air, but was laggier. At higher boost the 56 trim will become worth while though. The 20g is nearly maxed out at 25 psi in my example (most poeple run it at ~27-28 psi with boost falling off), while the 56 trim would be good for that 55 lbs/min at 30 psi... More headroom, and at that level it begins to counter the added lag with tons of extra power.
This is obviously a pretty involved topic, but I'll end it here for now before I get fired...