hmm . a 3.2 litre engine moves 1.6 litres per rotation. at 7000 rpm that's 11,200 litres / minute = 11.2 cubic meters per minute = 0.187 m3/s.
with w 74 mm ID TB, thats an area of 0.0043 square meters.
0.187 / 0.0043 = 43.5 metres/second (say 156 kmph - fyi, that's why ram air intakes don't work - you have to be going faster than 156 kmph to out push the suck, other aerodynamic inputs notwithstanding).
same analysis with the oem 67mm TB gets you 192 kmph.
i wouldn't call 156 kmph (100 mph) slow. i would say that the added work required to accelerate air (which is what the engine has to do - accelerate the air from zero to hero every time you blip the throttle) from 156 to 192 kmph is significant. this doesn't even take into consideration the friction losses associated with a smaller diameter tract. ps, may math could be wrong hey.
here's a good quote from a good article:
"The act of increasing the air velocity is not desirable when it does not contribute to the harmonic tuning of the intake pulses in the intake manifold. Since the intake duct is ahead of the throttle body and MAF sensor, these harmonics do not come into play."
Easy Performance | CAI System Design Consideration Primer
that is, we read about using a decreasing diameter intake to accelerate the air to facilitate cylinder fill at low rpms, and that opening up the intake track, while reducing friction at high rpm and increasing max hp, is detrimental to low rpm torque. however, note that most of this tuning is happening in the intake runners and resonance tube, and what is happening upstream of the throttle body is moot - just a supply of air, so best just ensure it gets there with as little friction and resistance as possible.
hijack! this is good discussion for video #1 of the series however, so i only feed kinda bad.