Poly1 is the usual CMOS gate electrode. It's going to be
there on any CMOS flow.
Poly2 is a cheap and easy way to get an isolated capacitor
for "mixed signal" applications. You can deposit or grow a
thin dielectric, lay down poly2, and there's your "POP cap".
This dielectric is meant to maximize areal capacitance,
while the poly-M1, M1-M2, M2-M3, ... dielectrics are
aimed at the oppposite (low interlayer capacitance) on
the one hand, but limited by things like via-filling, step
coverage and simple time-in-tube cost to a less than
infinite thickness.