Hi,
on my machine (pentium133/32MB Ram) "time" claims to have a 3 percent
shorter compilation time with make -j 2 than with just make.
Could this be true? I could imagine the machine doesnt have to 
wait so long for io operations to finish with two 
concurrent processes. Could this be the explanation?
In the same direction seems to lead the observation that 
the loadfactor always remains beyond 1.0 even if doing rather
big compilations.
ciao
Carsten