See below, from page 2:

You can get some pretty serious HP out of the J32, there are stroker kits out there for the NSX that'll push it to 3.6L. People have pushed almost 370whp out of the NSX.

Doesn't mean anything as even after all these years, VTEC and a good chassis still triumphs.

just posting because you're both largely (or partially) incorrect (in fact i signed up purely through frustration)

firstly, the smaller 'low rpm' turbo in the FD is designed purely so that it has better response and less of the lag

if you look at the dyno graph of a *well tuned* stock twin 13b you'll see it actually has REALLY flat torque, and the power is made very linearly to redline, much like an NA motor a la NSX..
stock tune the RX-7 isn't super smooth, big dip when the turbos cross over, but good tuning is good tuning

torque vs horsepower: here's where you are REALLY WRONG

torque IS horsepower, horsepower is simply torque over time (at an RPM) where torque is horsepower without time, the potential the engine has to push down at a given RPM

horsepower = torque (in ft-lb) x RPM /5252 (5252RPM is the point where torque and horsepower is identical, 500ft-lb at 5252rpm is 500 horsepower.)

so a motor with more low-down torque will make more power low down.. a motor with more torque up high will make more power up high than an engine without.

***at any given RPM, the motor that makes more torque will make more power*** - but yes it's true, if you have a low-torque motor (short stroke etc.) then revving higher, obviously (going by the above formula) will compensate in power production... hence formula 1 cars with half of nothing in the torque dept makes insane power.

toyota's VVTL is nothing NOTHING like VTEC, it's purely valve timing and lift in a very (relatively) minor way - more on this further on

the turbo's in the FD will also engage much lower than you're saying (4400rpm?) more like <2000 for the low boost smaller one.

and your quote: "let's just say the NSX would make 280HP CONSTANT between the 6500 and 8000 RPM" unlikely, it should have a FLAT torque curve (or flat-ish) with peak power production at or almost at redline. this means if it makes 300hp at 8000rpm, it makes 150hp at 4000rpm. this is obviously dependant on actual tuning, figures entirely pulled out of thin air.

ANY ENGINE PRODUCES POWER IN THE SAME MANNER. in fact, the FD has the potential to be tuned for a flat POWER curve and a tapering torque curve, as you can lower the boost as the revs rise, so the torque drops but the revs rise, giving a constant level of power.

and finally but most importantly... THE FD IS LIGHTER THAN THE NSX

a stock FD weighs ~ 1,150 kilograms (2,500 lb)
a stock NSX weighs 2,976 lb (1,350 kg) (91-92) 3,093lb (1,403kg)(93-97) 2,954lb (1,340kg) (02+)
2,809lb (1,274kg) (02+ NSX-R)

AT LEAST 300lb heavier is the LIGHTEST NSX.

more on vtec in next post:


ED:

whoops sorry 2,712 lb is the lightest NS-XR

yes, and the FD could have a bigger port and changed compression rotors, modified stock housing turbos to safely produce more boost etc. etc.

realistically, the FD in terms of power vs an NA NSX will always have the edge talking equivalent mods.

Kay kiddo. So since it's wikipedia, let's see where that particular section got its information from, shall we? Here we go...

...

...Funny! I don't see any references in that entire section. Now normally I'd just laugh in your face and tell you to get your facts straight, but I'm a generally intelligent fellow who has the bonus of being able to read Japanese, so I went and found an actual reliable source for that information:

http://www.honda.co.jp/news/1992/4921126.html

Which is a news posting Honda themselves put out back in 1992 announcing the Type R, stating that indeed, it's 1230 kg, 120 kg lighter than the 5MT NSX they had already put on the market. If Honda says that's what it is, then that's what it is. I hereby stand corrected and will edit my previous post.