well I tried to implement this pathway in MATLAB and it seems to me that it makes some sense. Here are two versions of the reaction progress simulation, one with low and the other with high affinity of the 3' to 5' exonuclease domain for the mismatches. The violet curve of the simulation corresponds the amount of synthesized dsDNA and the green curve corresponds to the error-free amount of that dsDNA.
The next figure shows that with low affinity of the exonuclease for the mismatches, the error rate is high (12%).
The figure below shows that the dsDNA synthesis error rate decreased with the increase of the affinity of the exonuclease for the mismatches.
The MATLAB code of this simulation is provided
Your comments and suggestions are most appreciated.