I'd also throw this into the mix:
The JWT is pre-tuned. Massive amounts of
dyno time has gone into every ECU JWT sells.
The SAFC can only properly be tuned on a dyno,
shooting for numbers based on other people's trial
and error on their own cars is basically shooting in
the dark, IMHO.
There is a third and 4th option that sounds kind
of promising. Similiar to the S-AFC, there is the
grassroots movement megasquirt opensource
project. You buy a kit, assemble it with some
amount of electronics soldering experience, and
run it with a base map to tune from. I'd consider
that to be one of the most difficult ways.
Then, there's going the Greddy way, buying an
e-manage and the appropriate map sensor, wiring
kit(s), and using a base map from a similiarly built
car and going to a dyno, spending a considerable
amount of money, time, and energy getting it
dialed in. Or, you could buy a wideband o2 and
maybe EGT gauge and try to get it dialed in
yourself.
It's easier to do on a chassis dyno, IMHO.
By the time you spend the money on an e-manage
setup, and the money on some dyno time, you may
or may not get the thing running as good as a JWT
ECU would. The nice thing is that if you order up
a JWT, and tell them exactly what MAF you have,
exactly what motor, injectors, and whatever other
variables they account for, like cams, the like, they
will send you an ECU programmed for good, solid
performance with out risking things like detonation.
As far as 'plug and play' goes, nothing beats JWT.
For the money, I'd probably go with the plug and
play. I doubt I want to travel for a couple hundred
miles to find a dyno shop that has a clue what they
are doing, and probably have to go back a few times
for 'adjustment' afterwards. JWT's already spent all
that time, on a lot of different configurations.
There, my $.02 goes a long-winded way, doesn't it?
Peter