Every time this emerges, I must break in with the minority view:
I regard R=V/I as simply the definition of the concept "resistance" (there is
also a "differential resistance" dV/dI) - in the same way as I regard C=Q/V and
L=-E/(dI/dt) as definitions of capacitance and inductance, where the concepts
apply (Kirchhoff situations?). Depending on construction and specific
circumstances the numerical values of R, C and L for various devices may or may
not be constants with respect to whatever variables you want to imagine. This
should cause no consternation or surprise.