I'm not sure, since we only do univariate polynomials and k-path has lots of variables, right?
But maybe this work can inspire looking for other small, constant factor saving circuits for different classes of polynomials. Would be cool!