Well that starts a whole new ball-game.
I was wondering (academic interest only) what you get if you port the heads, raise the compression ratio to about 10:1 and fit a more aggressive cam.
I did some research a while back correlating torque to compression ratio. I was trying at the time to put a torque number on a petrol industrial engine. The upshot was torque per litre is limited to around 10x your compression ratio.
8:1 and the best breathing engines struggle to break 80Nm per litre of displacement.
10:1 and the best breathing engines get 100Nm per litre of displacement.
The 9.35:1 3.9 v8 with no cats (source is a 1992 brochure for the "Brooklands" model) is 235 ft-lb (319Nm) which is ~82 Nm/litre at 2,600rpm and 185hp (137kw) at 4750rpm which is 286Nm. 90% of the peak torque.
These figures are a few percent better than the models with cats.
That suggests first up it's struggling to fill the cylinders (82Nm/litre when it could possibly be ~93Nm/litre) and could do with some port work.
If it were possible to get the 3.9V8 with port work and higher compression to 100Nm/litre then 390Nm at 2,500rpm and 90% of that at 4,750rpm would give you ~167kw.
So clearly more rpm is needed. If with a lumpier cam you can pull 80% of peak torque to 7000rpm then you've got ~230kw (310hp).
I don't think this is too far fetched, but it's an unlikely path to take when larger capacity (4.6 litre) is a bolt-in option, forced induction can do a lot more and a turbo diesel of the same age and capacity (4BD1T) can do that power with only external mods while getting almost twice the fuel economy.

