The resistor is there to make a limited current for your
"backstop" clamps, so that they can be smaller -and- keep gate
ox protected in the case that the front clamp is overvoltaged
at higher ESD currents (Vclamp+Iclamp1*Rclamp1 will be
held down to Vclamp by the second stage).
Thin film resistors have a finite power-to-blow and you want
to understand this limit, or you may have a "gate fuse"
instead. I have been surprised by short-pulse power-to-blow
data before, thermal calculations are poor predictors of how
the incident energy partitions into the resistor and the
surrounding materials. In the lab you would take a pulsed
overstress series looking for the beginning of resistor drift,
and stay below that power density (watts per square micron)
and energy density (joules per square micron) - both, or the
most restrictive / largest resistor area.
Diffused resistors can take more of a wallop, they are
contact limited rather than self-heating limited, much higher
volume / thermal mass.
However a diffused resistor used in the ESD network needs
to comply with latchup rules because it can be a trigger site,
forward substrate injection or reverse breakdown to well /
substrate can inject trigger current in a powered-ESD
scenario.