Recently, Dashen and Manohar have derived a set of consistency conditi
ons that must be satisfied by large-N(c) QCD, which enables them to gi
ve quite a general proof to that some of the large-N(c) spin-flavor sy
mmetry relations for baryon observables are violated only at order 1/N
(c)2. We try to clarify not only the power but also the limitation of
their model-independent approach through a comparative analysis of 1/N
(c) corrections to baryon observables in three simple effective models
of QCD. We then advocate a viewpoint that the N(c) --> infinity limit
and the 1/N(c) expansion framework should be clearly distinguished, b
ecause the former has little to do with our real world.